Available on x86-64 only.
Expand description
Platform-specific intrinsics for the x86_64
platform.
See the module documentation for more details.
Structs§
- Result of the
cpuid
instruction. - 128-bit wide set of four
f32
types, x86-specific - 256-bit wide set of eight
f32
types, x86-specific - 512-bit wide set of sixteen
f32
types, x86-specific - 128-bit wide set of two
f64
types, x86-specific - 128-bit wide integer vector type, x86-specific
- 256-bit wide set of four
f64
types, x86-specific - 256-bit wide integer vector type, x86-specific
- 512-bit wide set of eight
f64
types, x86-specific - 512-bit wide integer vector type, x86-specific
- __
m128bh Experimental 128-bit wide set of eightu16
types, x86-specific - __m128h
Experimental 128-bit wide set of 8f16
types, x86-specific - __
m256bh Experimental 256-bit wide set of 16u16
types, x86-specific - __m256h
Experimental 256-bit wide set of 16f16
types, x86-specific - __
m512bh Experimental 512-bit wide set of 32u16
types, x86-specific - __m512h
Experimental 512-bit wide set of 32f16
types, x86-specific - bf16
Experimental The BFloat16 type used in AVX-512 intrinsics.
Constants§
- Equal (ordered, non-signaling)
- Equal (ordered, signaling)
- Equal (unordered, non-signaling)
- Equal (unordered, signaling)
- False (ordered, non-signaling)
- False (ordered, signaling)
- Greater-than-or-equal (ordered, non-signaling)
- Greater-than-or-equal (ordered, signaling)
- Greater-than (ordered, non-signaling)
- Greater-than (ordered, signaling)
- Less-than-or-equal (ordered, non-signaling)
- Less-than-or-equal (ordered, signaling)
- Less-than (ordered, non-signaling)
- Less-than (ordered, signaling)
- Not-equal (ordered, non-signaling)
- Not-equal (ordered, signaling)
- Not-equal (unordered, non-signaling)
- Not-equal (unordered, signaling)
- Not-greater-than-or-equal (unordered, non-signaling)
- Not-greater-than-or-equal (unordered, signaling)
- Not-greater-than (unordered, non-signaling)
- Not-greater-than (unordered, signaling)
- Not-less-than-or-equal (unordered, non-signaling)
- Not-less-than-or-equal (unordered, signaling)
- Not-less-than (unordered, non-signaling)
- Not-less-than (unordered, signaling)
- Ordered (non-signaling)
- Ordered (signaling)
- True (unordered, non-signaling)
- True (unordered, signaling)
- Unordered (non-signaling)
- Unordered (signaling)
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- round up and do not suppress exceptions
- use MXCSR.RC; see
vendor::_MM_SET_ROUNDING_MODE
- round down and do not suppress exceptions
- use MXCSR.RC and suppress exceptions; see
vendor::_MM_SET_ROUNDING_MODE
- round to nearest and do not suppress exceptions
- suppress exceptions
- do not suppress exceptions
- use MXCSR.RC and do not suppress exceptions; see
vendor::_MM_SET_ROUNDING_MODE
- round to nearest
- round down
- round up
- truncate
- truncate and do not suppress exceptions
- See
_mm_prefetch
. - See
_mm_prefetch
. - See
_mm_prefetch
. - See
_mm_prefetch
. - See
_mm_prefetch
. - See
_mm_prefetch
. - See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- Mask only: return the bit mask
- For each character in
a
, find if it is inb
(Default) - The strings defined by
a
andb
are equal - Search for the defined substring in the target
- For each character in
a
, determine ifb[0] <= c <= b[1] or b[1] <= c <= b[2]...
- Index only: return the least significant bit (Default)
- Negates results only before the end of the string
- Do not negate results before the end of the string
- Index only: return the most significant bit
- Negates results
- Do not negate results (Default)
- String contains signed 8-bit characters
- String contains unsigned 16-bit characters
- String contains unsigned 8-bit characters (Default)
- Mask only: return the byte mask
- String contains unsigned 16-bit characters
XFEATURE_ENABLED_MASK
forXCR
- _MM_
CMPINT_ EQ Experimental Equal - _MM_
CMPINT_ FALSE Experimental False - _MM_
CMPINT_ LE Experimental Less-than-or-equal - _MM_
CMPINT_ LT Experimental Less-than - _MM_
CMPINT_ NE Experimental Not-equal - _MM_
CMPINT_ NLE Experimental Not less-than-or-equal - _MM_
CMPINT_ NLT Experimental Not less-than - _MM_
CMPINT_ TRUE Experimental True - _MM_
MANT_ NORM_ 1_ 2 Experimental interval [1, 2) - _MM_
MANT_ NORM_ P5_ 1 Experimental interval [0.5, 1) - _MM_
MANT_ NORM_ P5_ 2 Experimental interval [0.5, 2) - _MM_
MANT_ NORM_ P75_ 1P5 Experimental interval [0.75, 1.5) - _MM_
MANT_ SIGN_ NAN Experimental DEST = NaN if sign(SRC) = 1 - _MM_
MANT_ SIGN_ SRC Experimental sign = sign(SRC) - _MM_
MANT_ SIGN_ ZERO Experimental sign = 0 - _MM_
PERM_ AAAA Experimental - _MM_
PERM_ AAAB Experimental - _MM_
PERM_ AAAC Experimental - _MM_
PERM_ AAAD Experimental - _MM_
PERM_ AABA Experimental - _MM_
PERM_ AABB Experimental - _MM_
PERM_ AABC Experimental - _MM_
PERM_ AABD Experimental - _MM_
PERM_ AACA Experimental - _MM_
PERM_ AACB Experimental - _MM_
PERM_ AACC Experimental - _MM_
PERM_ AACD Experimental - _MM_
PERM_ AADA Experimental - _MM_
PERM_ AADB Experimental - _MM_
PERM_ AADC Experimental - _MM_
PERM_ AADD Experimental - _MM_
PERM_ ABAA Experimental - _MM_
PERM_ ABAB Experimental - _MM_
PERM_ ABAC Experimental - _MM_
PERM_ ABAD Experimental - _MM_
PERM_ ABBA Experimental - _MM_
PERM_ ABBB Experimental - _MM_
PERM_ ABBC Experimental - _MM_
PERM_ ABBD Experimental - _MM_
PERM_ ABCA Experimental - _MM_
PERM_ ABCB Experimental - _MM_
PERM_ ABCC Experimental - _MM_
PERM_ ABCD Experimental - _MM_
PERM_ ABDA Experimental - _MM_
PERM_ ABDB Experimental - _MM_
PERM_ ABDC Experimental - _MM_
PERM_ ABDD Experimental - _MM_
PERM_ ACAA Experimental - _MM_
PERM_ ACAB Experimental - _MM_
PERM_ ACAC Experimental - _MM_
PERM_ ACAD Experimental - _MM_
PERM_ ACBA Experimental - _MM_
PERM_ ACBB Experimental - _MM_
PERM_ ACBC Experimental - _MM_
PERM_ ACBD Experimental - _MM_
PERM_ ACCA Experimental - _MM_
PERM_ ACCB Experimental - _MM_
PERM_ ACCC Experimental - _MM_
PERM_ ACCD Experimental - _MM_
PERM_ ACDA Experimental - _MM_
PERM_ ACDB Experimental - _MM_
PERM_ ACDC Experimental - _MM_
PERM_ ACDD Experimental - _MM_
PERM_ ADAA Experimental - _MM_
PERM_ ADAB Experimental - _MM_
PERM_ ADAC Experimental - _MM_
PERM_ ADAD Experimental - _MM_
PERM_ ADBA Experimental - _MM_
PERM_ ADBB Experimental - _MM_
PERM_ ADBC Experimental - _MM_
PERM_ ADBD Experimental - _MM_
PERM_ ADCA Experimental - _MM_
PERM_ ADCB Experimental - _MM_
PERM_ ADCC Experimental - _MM_
PERM_ ADCD Experimental - _MM_
PERM_ ADDA Experimental - _MM_
PERM_ ADDB Experimental - _MM_
PERM_ ADDC Experimental - _MM_
PERM_ ADDD Experimental - _MM_
PERM_ BAAA Experimental - _MM_
PERM_ BAAB Experimental - _MM_
PERM_ BAAC Experimental - _MM_
PERM_ BAAD Experimental - _MM_
PERM_ BABA Experimental - _MM_
PERM_ BABB Experimental - _MM_
PERM_ BABC Experimental - _MM_
PERM_ BABD Experimental - _MM_
PERM_ BACA Experimental - _MM_
PERM_ BACB Experimental - _MM_
PERM_ BACC Experimental - _MM_
PERM_ BACD Experimental - _MM_
PERM_ BADA Experimental - _MM_
PERM_ BADB Experimental - _MM_
PERM_ BADC Experimental - _MM_
PERM_ BADD Experimental - _MM_
PERM_ BBAA Experimental - _MM_
PERM_ BBAB Experimental - _MM_
PERM_ BBAC Experimental - _MM_
PERM_ BBAD Experimental - _MM_
PERM_ BBBA Experimental - _MM_
PERM_ BBBB Experimental - _MM_
PERM_ BBBC Experimental - _MM_
PERM_ BBBD Experimental - _MM_
PERM_ BBCA Experimental - _MM_
PERM_ BBCB Experimental - _MM_
PERM_ BBCC Experimental - _MM_
PERM_ BBCD Experimental - _MM_
PERM_ BBDA Experimental - _MM_
PERM_ BBDB Experimental - _MM_
PERM_ BBDC Experimental - _MM_
PERM_ BBDD Experimental - _MM_
PERM_ BCAA Experimental - _MM_
PERM_ BCAB Experimental - _MM_
PERM_ BCAC Experimental - _MM_
PERM_ BCAD Experimental - _MM_
PERM_ BCBA Experimental - _MM_
PERM_ BCBB Experimental - _MM_
PERM_ BCBC Experimental - _MM_
PERM_ BCBD Experimental - _MM_
PERM_ BCCA Experimental - _MM_
PERM_ BCCB Experimental - _MM_
PERM_ BCCC Experimental - _MM_
PERM_ BCCD Experimental - _MM_
PERM_ BCDA Experimental - _MM_
PERM_ BCDB Experimental - _MM_
PERM_ BCDC Experimental - _MM_
PERM_ BCDD Experimental - _MM_
PERM_ BDAA Experimental - _MM_
PERM_ BDAB Experimental - _MM_
PERM_ BDAC Experimental - _MM_
PERM_ BDAD Experimental - _MM_
PERM_ BDBA Experimental - _MM_
PERM_ BDBB Experimental - _MM_
PERM_ BDBC Experimental - _MM_
PERM_ BDBD Experimental - _MM_
PERM_ BDCA Experimental - _MM_
PERM_ BDCB Experimental - _MM_
PERM_ BDCC Experimental - _MM_
PERM_ BDCD Experimental - _MM_
PERM_ BDDA Experimental - _MM_
PERM_ BDDB Experimental - _MM_
PERM_ BDDC Experimental - _MM_
PERM_ BDDD Experimental - _MM_
PERM_ CAAA Experimental - _MM_
PERM_ CAAB Experimental - _MM_
PERM_ CAAC Experimental - _MM_
PERM_ CAAD Experimental - _MM_
PERM_ CABA Experimental - _MM_
PERM_ CABB Experimental - _MM_
PERM_ CABC Experimental - _MM_
PERM_ CABD Experimental - _MM_
PERM_ CACA Experimental - _MM_
PERM_ CACB Experimental - _MM_
PERM_ CACC Experimental - _MM_
PERM_ CACD Experimental - _MM_
PERM_ CADA Experimental - _MM_
PERM_ CADB Experimental - _MM_
PERM_ CADC Experimental - _MM_
PERM_ CADD Experimental - _MM_
PERM_ CBAA Experimental - _MM_
PERM_ CBAB Experimental - _MM_
PERM_ CBAC Experimental - _MM_
PERM_ CBAD Experimental - _MM_
PERM_ CBBA Experimental - _MM_
PERM_ CBBB Experimental - _MM_
PERM_ CBBC Experimental - _MM_
PERM_ CBBD Experimental - _MM_
PERM_ CBCA Experimental - _MM_
PERM_ CBCB Experimental - _MM_
PERM_ CBCC Experimental - _MM_
PERM_ CBCD Experimental - _MM_
PERM_ CBDA Experimental - _MM_
PERM_ CBDB Experimental - _MM_
PERM_ CBDC Experimental - _MM_
PERM_ CBDD Experimental - _MM_
PERM_ CCAA Experimental - _MM_
PERM_ CCAB Experimental - _MM_
PERM_ CCAC Experimental - _MM_
PERM_ CCAD Experimental - _MM_
PERM_ CCBA Experimental - _MM_
PERM_ CCBB Experimental - _MM_
PERM_ CCBC Experimental - _MM_
PERM_ CCBD Experimental - _MM_
PERM_ CCCA Experimental - _MM_
PERM_ CCCB Experimental - _MM_
PERM_ CCCC Experimental - _MM_
PERM_ CCCD Experimental - _MM_
PERM_ CCDA Experimental - _MM_
PERM_ CCDB Experimental - _MM_
PERM_ CCDC Experimental - _MM_
PERM_ CCDD Experimental - _MM_
PERM_ CDAA Experimental - _MM_
PERM_ CDAB Experimental - _MM_
PERM_ CDAC Experimental - _MM_
PERM_ CDAD Experimental - _MM_
PERM_ CDBA Experimental - _MM_
PERM_ CDBB Experimental - _MM_
PERM_ CDBC Experimental - _MM_
PERM_ CDBD Experimental - _MM_
PERM_ CDCA Experimental - _MM_
PERM_ CDCB Experimental - _MM_
PERM_ CDCC Experimental - _MM_
PERM_ CDCD Experimental - _MM_
PERM_ CDDA Experimental - _MM_
PERM_ CDDB Experimental - _MM_
PERM_ CDDC Experimental - _MM_
PERM_ CDDD Experimental - _MM_
PERM_ DAAA Experimental - _MM_
PERM_ DAAB Experimental - _MM_
PERM_ DAAC Experimental - _MM_
PERM_ DAAD Experimental - _MM_
PERM_ DABA Experimental - _MM_
PERM_ DABB Experimental - _MM_
PERM_ DABC Experimental - _MM_
PERM_ DABD Experimental - _MM_
PERM_ DACA Experimental - _MM_
PERM_ DACB Experimental - _MM_
PERM_ DACC Experimental - _MM_
PERM_ DACD Experimental - _MM_
PERM_ DADA Experimental - _MM_
PERM_ DADB Experimental - _MM_
PERM_ DADC Experimental - _MM_
PERM_ DADD Experimental - _MM_
PERM_ DBAA Experimental - _MM_
PERM_ DBAB Experimental - _MM_
PERM_ DBAC Experimental - _MM_
PERM_ DBAD Experimental - _MM_
PERM_ DBBA Experimental - _MM_
PERM_ DBBB Experimental - _MM_
PERM_ DBBC Experimental - _MM_
PERM_ DBBD Experimental - _MM_
PERM_ DBCA Experimental - _MM_
PERM_ DBCB Experimental - _MM_
PERM_ DBCC Experimental - _MM_
PERM_ DBCD Experimental - _MM_
PERM_ DBDA Experimental - _MM_
PERM_ DBDB Experimental - _MM_
PERM_ DBDC Experimental - _MM_
PERM_ DBDD Experimental - _MM_
PERM_ DCAA Experimental - _MM_
PERM_ DCAB Experimental - _MM_
PERM_ DCAC Experimental - _MM_
PERM_ DCAD Experimental - _MM_
PERM_ DCBA Experimental - _MM_
PERM_ DCBB Experimental - _MM_
PERM_ DCBC Experimental - _MM_
PERM_ DCBD Experimental - _MM_
PERM_ DCCA Experimental - _MM_
PERM_ DCCB Experimental - _MM_
PERM_ DCCC Experimental - _MM_
PERM_ DCCD Experimental - _MM_
PERM_ DCDA Experimental - _MM_
PERM_ DCDB Experimental - _MM_
PERM_ DCDC Experimental - _MM_
PERM_ DCDD Experimental - _MM_
PERM_ DDAA Experimental - _MM_
PERM_ DDAB Experimental - _MM_
PERM_ DDAC Experimental - _MM_
PERM_ DDAD Experimental - _MM_
PERM_ DDBA Experimental - _MM_
PERM_ DDBB Experimental - _MM_
PERM_ DDBC Experimental - _MM_
PERM_ DDBD Experimental - _MM_
PERM_ DDCA Experimental - _MM_
PERM_ DDCB Experimental - _MM_
PERM_ DDCC Experimental - _MM_
PERM_ DDCD Experimental - _MM_
PERM_ DDDA Experimental - _MM_
PERM_ DDDB Experimental - _MM_
PERM_ DDDC Experimental - _MM_
PERM_ DDDD Experimental - _XABORT_
CAPACITY Experimental Transaction abort due to the transaction using too much memory. - _XABORT_
CONFLICT Experimental Transaction abort due to a memory conflict with another thread. - _XABORT_
DEBUG Experimental Transaction abort due to a debug trap. - _XABORT_
EXPLICIT Experimental Transaction explicitly aborted with xabort. The parameter passed to xabort is available with_xabort_code(status)
. - _XABORT_
NESTED Experimental Transaction abort in a inner nested transaction. - _XABORT_
RETRY Experimental Transaction retry is possible. - _XBEGIN_
STARTED Experimental Transaction successfully started.
Functions§
- _MM_
GET_ ⚠EXCEPTION_ MASK Deprecated See_mm_setcsr
- _MM_
GET_ ⚠EXCEPTION_ STATE Deprecated See_mm_setcsr
- _MM_
GET_ ⚠FLUSH_ ZERO_ MODE Deprecated See_mm_setcsr
- _MM_
GET_ ⚠ROUNDING_ MODE Deprecated See_mm_setcsr
- _MM_
SET_ ⚠EXCEPTION_ MASK Deprecated See_mm_setcsr
- _MM_
SET_ ⚠EXCEPTION_ STATE Deprecated See_mm_setcsr
- _MM_
SET_ ⚠FLUSH_ ZERO_ MODE Deprecated See_mm_setcsr
- _MM_
SET_ ⚠ROUNDING_ MODE Deprecated See_mm_setcsr
- Transpose the 4x4 matrix formed by 4 rows of __m128 in place.
- See
__cpuid_count
. - Returns the result of the
cpuid
instruction for a givenleaf
(EAX
) andsub_leaf
(ECX
). - Returns the highest-supported
leaf
(EAX
) and sub-leaf (ECX
)cpuid
values. - Reads the current value of the processor’s time-stamp counter and the
IA32_TSC_AUX MSR
. - Adds unsigned 32-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry or overflow flag), and store the unsigned 32-bit result inout
, and the carry-out is returned (carry or overflow flag). - Adds unsigned 64-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry or overflow flag), and store the unsigned 64-bit result inout
, and the carry-out is returned (carry or overflow flag). - Adds unsigned 32-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry or overflow flag), and store the unsigned 32-bit result inout
, and the carry-out is returned (carry or overflow flag). - Adds unsigned 64-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry or overflow flag), and store the unsigned 64-bit result inout
, and the carry-out is returned (carry or overflow flag). - Bitwise logical
AND
of inverteda
withb
. - Bitwise logical
AND
of inverteda
withb
. - Extracts bits of
a
specified bycontrol
into the least significant bits of the result. - Extracts bits of
a
specified bycontrol
into the least significant bits of the result. - Extracts bits in range [
start
,start
+length
) froma
into the least significant bits of the result. - Extracts bits in range [
start
,start
+length
) froma
into the least significant bits of the result. - Extracts bits of
a
specified bycontrol
into the least significant bits of the result. - Extracts bits of
a
specified bycontrol
into the least significant bits of the result. - Returns the bit in position
b
of the memory addressed byp
. - Returns the bit in position
b
of the memory addressed byp
. - Returns the bit in position
b
of the memory addressed byp
, then inverts that bit. - Returns the bit in position
b
of the memory addressed byp
, then inverts that bit. - Returns the bit in position
b
of the memory addressed byp
, then resets that bit to0
. - Returns the bit in position
b
of the memory addressed byp
, then resets that bit to0
. - Returns the bit in position
b
of the memory addressed byp
, then sets the bit to1
. - Returns the bit in position
b
of the memory addressed byp
, then sets the bit to1
. - Clears all bits below the least significant zero bit of
x
. - Clears all bits below the least significant zero bit of
x
. - Sets all bits of
x
to 1 except for the least significant zero bit. - Sets all bits of
x
to 1 except for the least significant zero bit. - Sets the least significant zero bit of
x
and clears all other bits. - Sets the least significant zero bit of
x
and clears all other bits. - Sets the least significant zero bit of
x
and clears all bits above that bit. - Sets the least significant zero bit of
x
and clears all bits above that bit. - Sets the least significant zero bit of
x
. - Sets the least significant zero bit of
x
. - Sets all bits of
x
below the least significant one. - Sets all bits of
x
below the least significant one. - Extracts lowest set isolated bit.
- Extracts lowest set isolated bit.
- Clears least significant bit and sets all other bits.
- Clears least significant bit and sets all other bits.
- Gets mask up to lowest set bit.
- Gets mask up to lowest set bit.
- Resets the lowest set bit of
x
. - Resets the lowest set bit of
x
. - Returns an integer with the reversed byte order of x
- Returns an integer with the reversed byte order of x
- Zeroes higher bits of
a
>=index
. - Zeroes higher bits of
a
>=index
. - Restores the
XMM
,MMX
,MXCSR
, andx87
FPU registers from the 512-byte-long 16-byte-aligned memory regionmem_addr
. - Restores the
XMM
,MMX
,MXCSR
, andx87
FPU registers from the 512-byte-long 16-byte-aligned memory regionmem_addr
. - Saves the
x87
FPU,MMX
technology,XMM
, andMXCSR
registers to the 512-byte-long 16-byte-aligned memory regionmem_addr
. - Saves the
x87
FPU,MMX
technology,XMM
, andMXCSR
registers to the 512-byte-long 16-byte-aligned memory regionmem_addr
. - Counts the leading most significant zero bits.
- Counts the leading most significant zero bits.
- Computes the absolute values of packed 8-bit integers in
a
. - Computes the absolute values of packed 16-bit integers in
a
. - Computes the absolute values of packed 32-bit integers in
a
. - Adds packed 8-bit integers in
a
andb
. - Adds packed 16-bit integers in
a
andb
. - Adds packed 32-bit integers in
a
andb
. - Adds packed 64-bit integers in
a
andb
. - Adds packed double-precision (64-bit) floating-point elements in
a
andb
. - Adds packed single-precision (32-bit) floating-point elements in
a
andb
. - Adds packed 8-bit integers in
a
andb
using saturation. - Adds packed 16-bit integers in
a
andb
using saturation. - Adds packed unsigned 8-bit integers in
a
andb
using saturation. - Adds packed unsigned 16-bit integers in
a
andb
using saturation. - Alternatively adds and subtracts packed double-precision (64-bit) floating-point elements in
a
to/from packed elements inb
. - Alternatively adds and subtracts packed single-precision (32-bit) floating-point elements in
a
to/from packed elements inb
. - Concatenates pairs of 16-byte blocks in
a
andb
into a 32-byte temporary result, shifts the result right byn
bytes, and returns the low 16 bytes. - Computes the bitwise AND of a packed double-precision (64-bit) floating-point elements in
a
andb
. - Computes the bitwise AND of packed single-precision (32-bit) floating-point elements in
a
andb
. - Computes the bitwise AND of 256 bits (representing integer data) in
a
andb
. - Computes the bitwise NOT of packed double-precision (64-bit) floating-point elements in
a
, and then AND withb
. - Computes the bitwise NOT of packed single-precision (32-bit) floating-point elements in
a
and then AND withb
. - Computes the bitwise NOT of 256 bits (representing integer data) in
a
and then AND withb
. - Averages packed unsigned 8-bit integers in
a
andb
. - Averages packed unsigned 16-bit integers in
a
andb
. - Blends packed 16-bit integers from
a
andb
using control maskIMM8
. - Blends packed 32-bit integers from
a
andb
using control maskIMM8
. - Blends packed double-precision (64-bit) floating-point elements from
a
andb
using control maskimm8
. - Blends packed single-precision (32-bit) floating-point elements from
a
andb
using control maskimm8
. - Blends packed 8-bit integers from
a
andb
usingmask
. - Blends packed double-precision (64-bit) floating-point elements from
a
andb
usingc
as a mask. - Blends packed single-precision (32-bit) floating-point elements from
a
andb
usingc
as a mask. - Broadcasts 128 bits from memory (composed of 2 packed double-precision (64-bit) floating-point elements) to all elements of the returned vector.
- Broadcasts 128 bits from memory (composed of 4 packed single-precision (32-bit) floating-point elements) to all elements of the returned vector.
- Broadcasts a double-precision (64-bit) floating-point element from memory to all elements of the returned vector.
- Broadcasts a single-precision (32-bit) floating-point element from memory to all elements of the returned vector.
- Broadcasts the low packed 8-bit integer from
a
to all elements of the 256-bit returned value. - Broadcasts the low packed 32-bit integer from
a
to all elements of the 256-bit returned value. - Broadcasts the low packed 64-bit integer from
a
to all elements of the 256-bit returned value. - Broadcasts the low double-precision (64-bit) floating-point element from
a
to all elements of the 256-bit returned value. - Broadcasts 128 bits of integer data from a to all 128-bit lanes in the 256-bit returned value.
- Broadcasts the low single-precision (32-bit) floating-point element from
a
to all elements of the 256-bit returned value. - Broadcasts the low packed 16-bit integer from a to all elements of the 256-bit returned value
- Shifts 128-bit lanes in
a
left byimm8
bytes while shifting in zeros. - Shifts 128-bit lanes in
a
right byimm8
bytes while shifting in zeros. - Casts vector of type __m128d to type __m256d; the upper 128 bits of the result are undefined.
- Casts vector of type __m256d to type __m128d.
- Cast vector of type __m256d to type __m256.
- Casts vector of type __m256d to type __m256i.
- Casts vector of type __m128 to type __m256; the upper 128 bits of the result are undefined.
- Casts vector of type __m256 to type __m128.
- Cast vector of type __m256 to type __m256d.
- Casts vector of type __m256 to type __m256i.
- Casts vector of type __m128i to type __m256i; the upper 128 bits of the result are undefined.
- Casts vector of type __m256i to type __m256d.
- Casts vector of type __m256i to type __m256.
- Casts vector of type __m256i to type __m128i.
- Rounds packed double-precision (64-bit) floating point elements in
a
toward positive infinity. - Rounds packed single-precision (32-bit) floating point elements in
a
toward positive infinity. - Compares packed double-precision (64-bit) floating-point elements in
a
andb
based on the comparison operand specified byIMM5
. - Compares packed single-precision (32-bit) floating-point elements in
a
andb
based on the comparison operand specified byIMM5
. - Compares packed 8-bit integers in
a
andb
for equality. - Compares packed 16-bit integers in
a
andb
for equality. - Compares packed 32-bit integers in
a
andb
for equality. - Compares packed 64-bit integers in
a
andb
for equality. - Compares packed 8-bit integers in
a
andb
for greater-than. - Compares packed 16-bit integers in
a
andb
for greater-than. - Compares packed 32-bit integers in
a
andb
for greater-than. - Compares packed 64-bit integers in
a
andb
for greater-than. - Sign-extend 8-bit integers to 16-bit integers.
- Sign-extend 8-bit integers to 32-bit integers.
- Sign-extend 8-bit integers to 64-bit integers.
- Sign-extend 16-bit integers to 32-bit integers.
- Sign-extend 16-bit integers to 64-bit integers.
- Sign-extend 32-bit integers to 64-bit integers.
- Converts packed 32-bit integers in
a
to packed double-precision (64-bit) floating-point elements. - Converts packed 32-bit integers in
a
to packed single-precision (32-bit) floating-point elements. - Zero-extend unsigned 8-bit integers in
a
to 16-bit integers. - Zero-extend the lower eight unsigned 8-bit integers in
a
to 32-bit integers. The upper eight elements ofa
are unused. - Zero-extend the lower four unsigned 8-bit integers in
a
to 64-bit integers. The upper twelve elements ofa
are unused. - Zeroes extend packed unsigned 16-bit integers in
a
to packed 32-bit integers, and stores the results indst
. - Zero-extend the lower four unsigned 16-bit integers in
a
to 64-bit integers. The upper four elements ofa
are unused. - Zero-extend unsigned 32-bit integers in
a
to 64-bit integers. - Converts packed double-precision (64-bit) floating-point elements in
a
to packed 32-bit integers. - Converts packed double-precision (64-bit) floating-point elements in
a
to packed single-precision (32-bit) floating-point elements. - Converts the 8 x 16-bit half-precision float values in the 128-bit vector
a
into 8 x 32-bit float values stored in a 256-bit wide vector. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed 32-bit integers. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed double-precision (64-bit) floating-point elements. - Converts the 8 x 32-bit float values in the 256-bit vector
a
into 8 x 16-bit half-precision float values stored in a 128-bit wide vector. - Returns the first element of the input vector of
[4 x double]
. - Returns the first element of the input vector of
[8 x i32]
. - Returns the first element of the input vector of
[8 x float]
. - Converts packed double-precision (64-bit) floating-point elements in
a
to packed 32-bit integers with truncation. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed 32-bit integers with truncation. - Computes the division of each of the 4 packed 64-bit floating-point elements in
a
by the corresponding packed elements inb
. - Computes the division of each of the 8 packed 32-bit floating-point elements in
a
by the corresponding packed elements inb
. - Conditionally multiplies the packed single-precision (32-bit) floating-point elements in
a
andb
using the high 4 bits inimm8
, sum the four products, and conditionally return the sum using the low 4 bits ofimm8
. - Extracts an 8-bit integer from
a
, selected withINDEX
. Returns a 32-bit integer containing the zero-extended integer data. - Extracts a 16-bit integer from
a
, selected withINDEX
. Returns a 32-bit integer containing the zero-extended integer data. - Extracts a 32-bit integer from
a
, selected withINDEX
. - Extracts a 64-bit integer from
a
, selected withINDEX
. - Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from
a
, selected withimm8
. - Extracts 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from
a
, selected withimm8
. - Extracts 128 bits (composed of integer data) from
a
, selected withimm8
. - Extracts 128 bits (of integer data) from
a
selected withIMM1
. - Rounds packed double-precision (64-bit) floating point elements in
a
toward negative infinity. - Rounds packed single-precision (32-bit) floating point elements in
a
toward negative infinity. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and add the intermediate result to packed elements inc
. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and add the intermediate result to packed elements inc
. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the intermediate result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and add the negated intermediate result to packed elements inc
. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and add the negated intermediate result to packed elements inc
. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. - Horizontally adds adjacent pairs of 16-bit integers in
a
andb
. - Horizontally adds adjacent pairs of 32-bit integers in
a
andb
. - Horizontal addition of adjacent pairs in the two packed vectors of 4 64-bit floating points
a
andb
. In the result, sums of elements froma
are returned in even locations, while sums of elements fromb
are returned in odd locations. - Horizontal addition of adjacent pairs in the two packed vectors of 8 32-bit floating points
a
andb
. In the result, sums of elements froma
are returned in locations of indices 0, 1, 4, 5; while sums of elements fromb
are locations 2, 3, 6, 7. - Horizontally adds adjacent pairs of 16-bit integers in
a
andb
using saturation. - Horizontally subtract adjacent pairs of 16-bit integers in
a
andb
. - Horizontally subtract adjacent pairs of 32-bit integers in
a
andb
. - Horizontal subtraction of adjacent pairs in the two packed vectors of 4 64-bit floating points
a
andb
. In the result, sums of elements froma
are returned in even locations, while sums of elements fromb
are returned in odd locations. - Horizontal subtraction of adjacent pairs in the two packed vectors of 8 32-bit floating points
a
andb
. In the result, sums of elements froma
are returned in locations of indices 0, 1, 4, 5; while sums of elements fromb
are locations 2, 3, 6, 7. - Horizontally subtract adjacent pairs of 16-bit integers in
a
andb
using saturation. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Copies
a
to result, and inserts the 8-bit integeri
into result at the location specified byindex
. - Copies
a
to result, and inserts the 16-bit integeri
into result at the location specified byindex
. - Copies
a
to result, and inserts the 32-bit integeri
into result at the location specified byindex
. - Copies
a
to result, and insert the 64-bit integeri
into result at the location specified byindex
. - Copies
a
to result, then inserts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) fromb
into result at the location specified byimm8
. - Copies
a
to result, then inserts 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) fromb
into result at the location specified byimm8
. - Copies
a
to result, then inserts 128 bits fromb
into result at the location specified byimm8
. - Copies
a
todst
, then insert 128 bits (of integer data) fromb
at the location specified byIMM1
. - Loads 256-bits of integer data from unaligned memory into result. This intrinsic may perform better than
_mm256_loadu_si256
when the data crosses a cache line boundary. - Loads 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) from memory into result.
mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Loads 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) from memory into result.
mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Loads 256-bits of integer data from memory into result.
mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Loads two 128-bit values (composed of 4 packed single-precision (32-bit) floating-point elements) from memory, and combine them into a 256-bit value.
hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Loads two 128-bit values (composed of 2 packed double-precision (64-bit) floating-point elements) from memory, and combine them into a 256-bit value.
hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Loads two 128-bit values (composed of integer data) from memory, and combine them into a 256-bit value.
hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Loads 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) from memory into result.
mem_addr
does not need to be aligned on any particular boundary. - Loads 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) from memory into result.
mem_addr
does not need to be aligned on any particular boundary. - Loads 256-bits of integer data from memory into result.
mem_addr
does not need to be aligned on any particular boundary. - Multiplies packed signed 16-bit integers in
a
andb
, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers. - Vertically multiplies each unsigned 8-bit integer from
a
with the corresponding signed 8-bit integer fromb
, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Loads packed 32-bit integers from memory pointed by
mem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - Loads packed 64-bit integers from memory pointed by
mem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - Loads packed double-precision (64-bit) floating-point elements from memory into result using
mask
(elements are zeroed out when the high bit of the corresponding element is not set). - Loads packed single-precision (32-bit) floating-point elements from memory into result using
mask
(elements are zeroed out when the high bit of the corresponding element is not set). - Stores packed 32-bit integers from
a
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - Stores packed 64-bit integers from
a
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - Stores packed double-precision (64-bit) floating-point elements from
a
into memory usingmask
. - Stores packed single-precision (32-bit) floating-point elements from
a
into memory usingmask
. - Compares packed 8-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed 16-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed 32-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed unsigned 8-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed unsigned 16-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed unsigned 32-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed double-precision (64-bit) floating-point elements in
a
andb
, and returns packed maximum values - Compares packed single-precision (32-bit) floating-point elements in
a
andb
, and returns packed maximum values - Compares packed 8-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed 16-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed 32-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed unsigned 8-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed unsigned 16-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed unsigned 32-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed double-precision (64-bit) floating-point elements in
a
andb
, and returns packed minimum values - Compares packed single-precision (32-bit) floating-point elements in
a
andb
, and returns packed minimum values - Duplicate even-indexed double-precision (64-bit) floating-point elements from
a
, and returns the results. - Duplicate odd-indexed single-precision (32-bit) floating-point elements from
a
, and returns the results. - Duplicate even-indexed single-precision (32-bit) floating-point elements from
a
, and returns the results. - Creates mask from the most significant bit of each 8-bit element in
a
, return the result. - Sets each bit of the returned mask based on the most significant bit of the corresponding packed double-precision (64-bit) floating-point element in
a
. - Sets each bit of the returned mask based on the most significant bit of the corresponding packed single-precision (32-bit) floating-point element in
a
. - Computes the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in
a
compared to those inb
, and stores the 16-bit results in dst. Eight SADs are performed for each 128-bit lane using one quadruplet fromb
and eight quadruplets froma
. One quadruplet is selected fromb
starting at on the offset specified inimm8
. Eight quadruplets are formed from sequential 8-bit integers selected froma
starting at the offset specified inimm8
. - Multiplies the low 32-bit integers from each packed 64-bit element in
a
andb
- Multiplies the low unsigned 32-bit integers from each packed 64-bit element in
a
andb
- Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
. - Multiplies the packed 16-bit integers in
a
andb
, producing intermediate 32-bit integers and returning the high 16 bits of the intermediate integers. - Multiplies the packed unsigned 16-bit integers in
a
andb
, producing intermediate 32-bit integers and returning the high 16 bits of the intermediate integers. - Multiplies packed 16-bit integers in
a
andb
, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and return bits[16:1]
. - Multiplies the packed 16-bit integers in
a
andb
, producing intermediate 32-bit integers, and returns the low 16 bits of the intermediate integers - Multiplies the packed 32-bit integers in
a
andb
, producing intermediate 64-bit integers, and returns the low 32 bits of the intermediate integers - Computes the bitwise OR packed double-precision (64-bit) floating-point elements in
a
andb
. - Computes the bitwise OR packed single-precision (32-bit) floating-point elements in
a
andb
. - Computes the bitwise OR of 256 bits (representing integer data) in
a
andb
- Converts packed 16-bit integers from
a
andb
to packed 8-bit integers using signed saturation - Converts packed 32-bit integers from
a
andb
to packed 16-bit integers using signed saturation - Converts packed 16-bit integers from
a
andb
to packed 8-bit integers using unsigned saturation - Converts packed 32-bit integers from
a
andb
to packed 16-bit integers using unsigned saturation - Shuffles 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) selected by
imm8
froma
andb
. - Shuffles 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) selected by
imm8
froma
andb
. - Shuffles 128-bits (composed of integer data) selected by
imm8
froma
andb
. - Shuffles 128-bits of integer data selected by
imm8
froma
andb
. - Permutes 64-bit integers from
a
using control maskimm8
. - Shuffles 64-bit floating-point elements in
a
across lanes using the control inimm8
. - Shuffles double-precision (64-bit) floating-point elements in
a
within 128-bit lanes using the control inimm8
. - Shuffles single-precision (32-bit) floating-point elements in
a
within 128-bit lanes using the control inimm8
. - Permutes packed 32-bit integers from
a
according to the content ofb
. - Shuffles eight 32-bit floating-point elements in
a
across lanes using the corresponding 32-bit integer index inidx
. - Shuffles double-precision (64-bit) floating-point elements in
a
within 256-bit lanes using the control inb
. - Shuffles single-precision (32-bit) floating-point elements in
a
within 128-bit lanes using the control inb
. - Computes the approximate reciprocal of packed single-precision (32-bit) floating-point elements in
a
, and returns the results. The maximum relative error for this approximation is less than 1.5*2^-12. - Rounds packed double-precision (64-bit) floating point elements in
a
according to the flagROUNDING
. The value ofROUNDING
may be as follows: - Rounds packed single-precision (32-bit) floating point elements in
a
according to the flagROUNDING
. The value ofROUNDING
may be as follows: - Computes the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in
a
, and returns the results. The maximum relative error for this approximation is less than 1.5*2^-12. - Computes the absolute differences of packed unsigned 8-bit integers in
a
andb
, then horizontally sum each consecutive 8 differences to produce four unsigned 16-bit integers, and pack these unsigned 16-bit integers in the low 16 bits of the 64-bit return value - Broadcasts 8-bit integer
a
to all elements of returned vector. This intrinsic may generate thevpbroadcastb
. - Broadcasts 16-bit integer
a
to all elements of returned vector. This intrinsic may generate thevpbroadcastw
. - Broadcasts 32-bit integer
a
to all elements of returned vector. This intrinsic may generate thevpbroadcastd
. - Broadcasts 64-bit integer
a
to all elements of returned vector. This intrinsic may generate thevpbroadcastq
. - Broadcasts double-precision (64-bit) floating-point value
a
to all elements of returned vector. - Broadcasts single-precision (32-bit) floating-point value
a
to all elements of returned vector. - Sets packed 8-bit integers in returned vector with the supplied values.
- Sets packed 16-bit integers in returned vector with the supplied values.
- Sets packed 32-bit integers in returned vector with the supplied values.
- Sets packed 64-bit integers in returned vector with the supplied values.
- Sets packed __m256 returned vector with the supplied values.
- Sets packed __m256d returned vector with the supplied values.
- Sets packed __m256i returned vector with the supplied values.
- Sets packed double-precision (64-bit) floating-point elements in returned vector with the supplied values.
- Sets packed single-precision (32-bit) floating-point elements in returned vector with the supplied values.
- Sets packed 8-bit integers in returned vector with the supplied values in reverse order.
- Sets packed 16-bit integers in returned vector with the supplied values in reverse order.
- Sets packed 32-bit integers in returned vector with the supplied values in reverse order.
- Sets packed 64-bit integers in returned vector with the supplied values in reverse order.
- Sets packed __m256 returned vector with the supplied values.
- Sets packed __m256d returned vector with the supplied values.
- Sets packed __m256i returned vector with the supplied values.
- Sets packed double-precision (64-bit) floating-point elements in returned vector with the supplied values in reverse order.
- Sets packed single-precision (32-bit) floating-point elements in returned vector with the supplied values in reverse order.
- Returns vector of type __m256d with all elements set to zero.
- Returns vector of type __m256 with all elements set to zero.
- Returns vector of type __m256i with all elements set to zero.
- Shuffles bytes from
a
according to the content ofb
. - Shuffles 32-bit integers in 128-bit lanes of
a
using the control inimm8
. - Shuffles double-precision (64-bit) floating-point elements within 128-bit lanes using the control in
imm8
. - Shuffles single-precision (32-bit) floating-point elements in
a
within 128-bit lanes using the control inimm8
. - Shuffles 16-bit integers in the high 64 bits of 128-bit lanes of
a
using the control inimm8
. The low 64 bits of 128-bit lanes ofa
are copied to the output. - Shuffles 16-bit integers in the low 64 bits of 128-bit lanes of
a
using the control inimm8
. The high 64 bits of 128-bit lanes ofa
are copied to the output. - Negates packed 8-bit integers in
a
when the corresponding signed 8-bit integer inb
is negative, and returns the results. Results are zeroed out when the corresponding element inb
is zero. - Negates packed 16-bit integers in
a
when the corresponding signed 16-bit integer inb
is negative, and returns the results. Results are zeroed out when the corresponding element inb
is zero. - Negates packed 32-bit integers in
a
when the corresponding signed 32-bit integer inb
is negative, and returns the results. Results are zeroed out when the corresponding element inb
is zero. - Shifts packed 16-bit integers in
a
left bycount
while shifting in zeros, and returns the result - Shifts packed 32-bit integers in
a
left bycount
while shifting in zeros, and returns the result - Shifts packed 64-bit integers in
a
left bycount
while shifting in zeros, and returns the result - Shifts packed 16-bit integers in
a
left byIMM8
while shifting in zeros, return the results; - Shifts packed 32-bit integers in
a
left byIMM8
while shifting in zeros, return the results; - Shifts packed 64-bit integers in
a
left byIMM8
while shifting in zeros, return the results; - Shifts 128-bit lanes in
a
left byimm8
bytes while shifting in zeros. - Shifts packed 32-bit integers in
a
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - Shifts packed 64-bit integers in
a
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - Returns the square root of packed double-precision (64-bit) floating point elements in
a
. - Returns the square root of packed single-precision (32-bit) floating point elements in
a
. - Shifts packed 16-bit integers in
a
right bycount
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right bycount
while shifting in sign bits. - Shifts packed 16-bit integers in
a
right byIMM8
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right byIMM8
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in sign bits. - Shifts packed 16-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 32-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 64-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 16-bit integers in
a
right byIMM8
while shifting in zeros - Shifts packed 32-bit integers in
a
right byIMM8
while shifting in zeros - Shifts packed 64-bit integers in
a
right byIMM8
while shifting in zeros - Shifts 128-bit lanes in
a
right byimm8
bytes while shifting in zeros. - Shifts packed 32-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in zeros, - Shifts packed 64-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in zeros, - Stores 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) from
a
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Stores 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) from
a
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Stores 256-bits of integer data from
a
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - Stores the high and low 128-bit halves (each composed of 4 packed single-precision (32-bit) floating-point elements) from
a
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Stores the high and low 128-bit halves (each composed of 2 packed double-precision (64-bit) floating-point elements) from
a
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Stores the high and low 128-bit halves (each composed of integer data) from
a
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - Stores 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) from
a
into memory.mem_addr
does not need to be aligned on any particular boundary. - Stores 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) from
a
into memory.mem_addr
does not need to be aligned on any particular boundary. - Stores 256-bits of integer data from
a
into memory.mem_addr
does not need to be aligned on any particular boundary. - Load 256-bits of integer data from memory into dst using a non-temporal memory hint. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon)
- Moves double-precision values from a 256-bit vector of
[4 x double]
to a 32-byte aligned memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon). - Moves single-precision floating point values from a 256-bit vector of
[8 x float]
to a 32-byte aligned memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon). - Moves integer data from a 256-bit integer vector to a 32-byte aligned memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon)
- Subtract packed 8-bit integers in
b
from packed 8-bit integers ina
- Subtract packed 16-bit integers in
b
from packed 16-bit integers ina
- Subtract packed 32-bit integers in
b
from packed 32-bit integers ina
- Subtract packed 64-bit integers in
b
from packed 64-bit integers ina
- Subtracts packed double-precision (64-bit) floating-point elements in
b
from packed elements ina
. - Subtracts packed single-precision (32-bit) floating-point elements in
b
from packed elements ina
. - Subtract packed 8-bit integers in
b
from packed 8-bit integers ina
using saturation. - Subtract packed 16-bit integers in
b
from packed 16-bit integers ina
using saturation. - Subtract packed unsigned 8-bit integers in
b
from packed 8-bit integers ina
using saturation. - Subtract packed unsigned 16-bit integers in
b
from packed 16-bit integers ina
using saturation. - Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theCF
value. - Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theCF
value. - Computes the bitwise AND of 256 bits (representing integer data) in
a
andb
, and setZF
to 1 if the result is zero, otherwise setZF
to 0. Computes the bitwise NOT ofa
and then AND withb
, and setCF
to 1 if the result is zero, otherwise setCF
to 0. Return theCF
value. - Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return 1 if both theZF
andCF
values are zero, otherwise return 0. - Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return 1 if both theZF
andCF
values are zero, otherwise return 0. - Computes the bitwise AND of 256 bits (representing integer data) in
a
andb
, and setZF
to 1 if the result is zero, otherwise setZF
to 0. Computes the bitwise NOT ofa
and then AND withb
, and setCF
to 1 if the result is zero, otherwise setCF
to 0. Return 1 if both theZF
andCF
values are zero, otherwise return 0. - Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theZF
value. - Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 256-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theZF
value. - Computes the bitwise AND of 256 bits (representing integer data) in
a
andb
, and setZF
to 1 if the result is zero, otherwise setZF
to 0. Computes the bitwise NOT ofa
and then AND withb
, and setCF
to 1 if the result is zero, otherwise setCF
to 0. Return theZF
value. - Returns vector of type
__m256d
with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Returns vector of type
__m256
with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Returns vector of type __m256i with with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent to
mem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Unpacks and interleave 8-bit integers from the high half of each 128-bit lane in
a
andb
. - Unpacks and interleave 16-bit integers from the high half of each 128-bit lane of
a
andb
. - Unpacks and interleave 32-bit integers from the high half of each 128-bit lane of
a
andb
. - Unpacks and interleave 64-bit integers from the high half of each 128-bit lane of
a
andb
. - Unpacks and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in
a
andb
. - Unpacks and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in
a
andb
. - Unpacks and interleave 8-bit integers from the low half of each 128-bit lane of
a
andb
. - Unpacks and interleave 16-bit integers from the low half of each 128-bit lane of
a
andb
. - Unpacks and interleave 32-bit integers from the low half of each 128-bit lane of
a
andb
. - Unpacks and interleave 64-bit integers from the low half of each 128-bit lane of
a
andb
. - Unpacks and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in
a
andb
. - Unpacks and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in
a
andb
. - Computes the bitwise XOR of packed double-precision (64-bit) floating-point elements in
a
andb
. - Computes the bitwise XOR of packed single-precision (32-bit) floating-point elements in
a
andb
. - Computes the bitwise XOR of 256 bits (representing integer data) in
a
andb
- Zeroes the contents of all XMM or YMM registers.
- Zeroes the upper 128 bits of all YMM registers; the lower 128-bits of the registers are unmodified.
- Constructs a 256-bit floating-point vector of
[4 x double]
from a 128-bit floating-point vector of[2 x double]
. The lower 128 bits contain the value of the source vector. The upper 128 bits are set to zero. - Constructs a 256-bit floating-point vector of
[8 x float]
from a 128-bit floating-point vector of[4 x float]
. The lower 128 bits contain the value of the source vector. The upper 128 bits are set to zero. - Constructs a 256-bit integer vector from a 128-bit integer vector. The lower 128 bits contain the value of the source vector. The upper 128 bits are set to zero.
- Computes the absolute value of packed 8-bit signed integers in
a
and return the unsigned results. - Computes the absolute value of each of the packed 16-bit signed integers in
a
and return the 16-bit unsigned integer - Computes the absolute value of each of the packed 32-bit signed integers in
a
and return the 32-bit unsigned integer - Adds packed 8-bit integers in
a
andb
. - Adds packed 16-bit integers in
a
andb
. - Adds packed 32-bit integers in
a
andb
. - Adds packed 64-bit integers in
a
andb
. - Adds packed double-precision (64-bit) floating-point elements in
a
andb
. - Adds __m128 vectors.
- Returns a new vector with the low element of
a
replaced by the sum of the low elements ofa
andb
. - Adds the first component of
a
andb
, the other components are copied froma
. - Adds packed 8-bit integers in
a
andb
using saturation. - Adds packed 16-bit integers in
a
andb
using saturation. - Adds packed unsigned 8-bit integers in
a
andb
using saturation. - Adds packed unsigned 16-bit integers in
a
andb
using saturation. - Alternatively add and subtract packed double-precision (64-bit) floating-point elements in
a
to/from packed elements inb
. - Alternatively add and subtract packed single-precision (32-bit) floating-point elements in
a
to/from packed elements inb
. - Performs one round of an AES decryption flow on data (state) in
a
. - Performs the last round of an AES decryption flow on data (state) in
a
. - Performs one round of an AES encryption flow on data (state) in
a
. - Performs the last round of an AES encryption flow on data (state) in
a
. - Performs the
InvMixColumns
transformation ona
. - Assist in expanding the AES cipher key.
- Concatenate 16-byte blocks in
a
andb
into a 32-byte temporary result, shift the result right byn
bytes, and returns the low 16 bytes. - Computes the bitwise AND of packed double-precision (64-bit) floating-point elements in
a
andb
. - Bitwise AND of packed single-precision (32-bit) floating-point elements.
- Computes the bitwise AND of 128 bits (representing integer data) in
a
andb
. - Computes the bitwise NOT of
a
and then AND withb
. - Bitwise AND-NOT of packed single-precision (32-bit) floating-point elements.
- Computes the bitwise NOT of 128 bits (representing integer data) in
a
and then AND withb
. - Averages packed unsigned 8-bit integers in
a
andb
. - Averages packed unsigned 16-bit integers in
a
andb
. - Blend packed 16-bit integers from
a
andb
using the maskIMM8
. - Blends packed 32-bit integers from
a
andb
using control maskIMM4
. - Blend packed double-precision (64-bit) floating-point elements from
a
andb
using control maskIMM2
- Blend packed single-precision (32-bit) floating-point elements from
a
andb
using maskIMM4
- Blend packed 8-bit integers from
a
andb
usingmask
- Blend packed double-precision (64-bit) floating-point elements from
a
andb
usingmask
- Blend packed single-precision (32-bit) floating-point elements from
a
andb
usingmask
- Broadcasts a single-precision (32-bit) floating-point element from memory to all elements of the returned vector.
- Broadcasts the low packed 8-bit integer from
a
to all elements of the 128-bit returned value. - Broadcasts the low packed 32-bit integer from
a
to all elements of the 128-bit returned value. - Broadcasts the low packed 64-bit integer from
a
to all elements of the 128-bit returned value. - Broadcasts the low double-precision (64-bit) floating-point element from
a
to all elements of the 128-bit returned value. - Broadcasts 128 bits of integer data from a to all 128-bit lanes in the 256-bit returned value.
- Broadcasts the low single-precision (32-bit) floating-point element from
a
to all elements of the 128-bit returned value. - Broadcasts the low packed 16-bit integer from a to all elements of the 128-bit returned value
- Shifts
a
left byIMM8
bytes while shifting in zeros. - Shifts
a
right byIMM8
bytes while shifting in zeros. - Casts a 128-bit floating-point vector of
[2 x double]
into a 128-bit floating-point vector of[4 x float]
. - Casts a 128-bit floating-point vector of
[2 x double]
into a 128-bit integer vector. - Casts a 128-bit floating-point vector of
[4 x float]
into a 128-bit floating-point vector of[2 x double]
. - Casts a 128-bit floating-point vector of
[4 x float]
into a 128-bit integer vector. - Casts a 128-bit integer vector into a 128-bit floating-point vector of
[2 x double]
. - Casts a 128-bit integer vector into a 128-bit floating-point vector of
[4 x float]
. - Round the packed double-precision (64-bit) floating-point elements in
a
up to an integer value, and stores the results as packed double-precision floating-point elements. - Round the packed single-precision (32-bit) floating-point elements in
a
up to an integer value, and stores the results as packed single-precision floating-point elements. - Round the lower double-precision (64-bit) floating-point element in
b
up to an integer value, store the result as a double-precision floating-point element in the lower element of the intrinsic result, and copies the upper element froma
to the upper element of the intrinsic result. - Round the lower single-precision (32-bit) floating-point element in
b
up to an integer value, store the result as a single-precision floating-point element in the lower element of the intrinsic result, and copies the upper 3 packed elements froma
to the upper elements of the intrinsic result. - Invalidates and flushes the cache line that contains
p
from all levels of the cache hierarchy. - Performs a carry-less multiplication of two 64-bit polynomials over the finite field GF(2).
- Compares packed double-precision (64-bit) floating-point elements in
a
andb
based on the comparison operand specified byIMM5
. - Compares packed single-precision (32-bit) floating-point elements in
a
andb
based on the comparison operand specified byIMM5
. - Compares the lower double-precision (64-bit) floating-point element in
a
andb
based on the comparison operand specified byIMM5
, store the result in the lower element of returned vector, and copies the upper element froma
to the upper element of returned vector. - Compares the lower single-precision (32-bit) floating-point element in
a
andb
based on the comparison operand specified byIMM5
, store the result in the lower element of returned vector, and copies the upper 3 packed elements froma
to the upper elements of returned vector. - Compares packed 8-bit integers in
a
andb
for equality. - Compares packed 16-bit integers in
a
andb
for equality. - Compares packed 32-bit integers in
a
andb
for equality. - Compares packed 64-bit integers in
a
andb
for equality - Compares corresponding elements in
a
andb
for equality. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input elements were equal, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the equality comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for equality. The lowest 32 bits of the result will be0xffffffff
if the two inputs are equal, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return1
ifb
did not contain a null character and the resulting mask was zero, and0
otherwise. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if the resulting mask was non-zero, and0
otherwise. - Compares packed strings
a
andb
with lengthsla
andlb
using the control inIMM8
and return the generated index. Similar to_mm_cmpistri
with the exception that_mm_cmpistri
implicitly determines the length ofa
andb
. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return the generated mask. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return bit0
of the resulting bit mask. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if any character in a was null, and0
otherwise. - Compares packed strings in
a
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if any character inb
was null, and0
otherwise. - Compares corresponding elements in
a
andb
for greater-than-or-equal. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is greater than or equal to the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the greater-than-or-equal comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for greater than or equal. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is greater than or equalb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares packed 8-bit integers in
a
andb
for greater-than. - Compares packed 16-bit integers in
a
andb
for greater-than. - Compares packed 32-bit integers in
a
andb
for greater-than. - Compares packed 64-bit integers in
a
andb
for greater-than, return the results. - Compares corresponding elements in
a
andb
for greater-than. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is greater than the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the greater-than comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for greater than. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is greater thanb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and return1
ifb
did not contain a null character and the resulting mask was zero, and0
otherwise. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and return1
if the resulting mask was non-zero, and0
otherwise. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
and return the generated index. Similar to_mm_cmpestri
with the exception that_mm_cmpestri
requires the lengths ofa
andb
to be explicitly specified. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and return the generated mask. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and return bit0
of the resulting bit mask. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and returns1
if any character ina
was null, and0
otherwise. - Compares packed strings with implicit lengths in
a
andb
using the control inIMM8
, and return1
if any character inb
was null. and0
otherwise. - Compares corresponding elements in
a
andb
for less-than-or-equal - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is less than or equal to the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the less-than-or-equal comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for less than or equal. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is less than or equalb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares packed 8-bit integers in
a
andb
for less-than. - Compares packed 16-bit integers in
a
andb
for less-than. - Compares packed 32-bit integers in
a
andb
for less-than. - Compares corresponding elements in
a
andb
for less-than. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is less than the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the less-than comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for less than. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is less thanb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
for not-equal. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input elements are not equal, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the not-equal comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for inequality. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is not equal tob.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
for not-greater-than-or-equal. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is not greater than or equal to the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the not-greater-than-or-equal comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for not-greater-than-or-equal. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is not greater than or equal tob.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
for not-greater-than. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is not greater than the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the not-greater-than comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for not-greater-than. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is not greater thanb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
for not-less-than-or-equal. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is not less than or equal to the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the not-less-than-or-equal comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for not-less-than-or-equal. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is not less than or equal tob.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
for not-less-than. - Compares each of the four floats in
a
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input element ina
is not less than the corresponding element inb
, or0
otherwise. - Returns a new vector with the low element of
a
replaced by the not-less-than comparison of the lower elements ofa
andb
. - Compares the lowest
f32
of both inputs for not-less-than. The lowest 32 bits of the result will be0xffffffff
ifa.extract(0)
is not less thanb.extract(0)
, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
to see if neither isNaN
. - Compares each of the four floats in
a
to the corresponding element inb
. Returns four floats that have one of two possible bit patterns. The element in the output vector will be0xffffffff
if the input elements ina
andb
are ordered (i.e., neither of them is a NaN), or 0 otherwise. - Returns a new vector with the low element of
a
replaced by the result of comparing both of the lower elements ofa
andb
toNaN
. If neither are equal toNaN
then0xFFFFFFFFFFFFFFFF
is used and0
otherwise. - Checks if the lowest
f32
of both inputs are ordered. The lowest 32 bits of the result will be0xffffffff
if neither ofa.extract(0)
orb.extract(0)
is a NaN, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares corresponding elements in
a
andb
to see if either isNaN
. - Compares each of the four floats in
a
to the corresponding element inb
. Returns four floats that have one of two possible bit patterns. The element in the output vector will be0xffffffff
if the input elements ina
andb
are unordered (i.e., at least on of them is a NaN), or 0 otherwise. - Returns a new vector with the low element of
a
replaced by the result of comparing both of the lower elements ofa
andb
toNaN
. If either is equal toNaN
then0xFFFFFFFFFFFFFFFF
is used and0
otherwise. - Checks if the lowest
f32
of both inputs are unordered. The lowest 32 bits of the result will be0xffffffff
if any ofa.extract(0)
orb.extract(0)
is a NaN, or0
otherwise. The upper 96 bits of the result are the upper 96 bits ofa
. - Compares the lower element of
a
andb
for equality. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if they are equal, or0
otherwise. - Compares the lower element of
a
andb
for greater-than-or-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is greater than or equal to the one fromb
, or0
otherwise. - Compares the lower element of
a
andb
for greater-than. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is greater than the one fromb
, or0
otherwise. - Compares the lower element of
a
andb
for less-than-or-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is less than or equal to the one fromb
, or0
otherwise. - Compares the lower element of
a
andb
for less-than. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is less than the one fromb
, or0
otherwise. - Compares the lower element of
a
andb
for not-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if they are not equal, or0
otherwise. - Starting with the initial value in
crc
, return the accumulated CRC32-C value for unsigned 8-bit integerv
. - Starting with the initial value in
crc
, return the accumulated CRC32-C value for unsigned 16-bit integerv
. - Starting with the initial value in
crc
, return the accumulated CRC32-C value for unsigned 32-bit integerv
. - Starting with the initial value in
crc
, return the accumulated CRC32-C value for unsigned 64-bit integerv
. - Alias for
_mm_cvtsi32_ss
. - Alias for
_mm_cvtss_si32
. - Sign extend packed 8-bit integers in
a
to packed 16-bit integers - Sign extend packed 8-bit integers in
a
to packed 32-bit integers - Sign extend packed 8-bit integers in the low 8 bytes of
a
to packed 64-bit integers - Sign extend packed 16-bit integers in
a
to packed 32-bit integers - Sign extend packed 16-bit integers in
a
to packed 64-bit integers - Sign extend packed 32-bit integers in
a
to packed 64-bit integers - Converts the lower two packed 32-bit integers in
a
to packed double-precision (64-bit) floating-point elements. - Converts packed 32-bit integers in
a
to packed single-precision (32-bit) floating-point elements. - Zeroes extend packed unsigned 8-bit integers in
a
to packed 16-bit integers - Zeroes extend packed unsigned 8-bit integers in
a
to packed 32-bit integers - Zeroes extend packed unsigned 8-bit integers in
a
to packed 64-bit integers - Zeroes extend packed unsigned 16-bit integers in
a
to packed 32-bit integers - Zeroes extend packed unsigned 16-bit integers in
a
to packed 64-bit integers - Zeroes extend packed unsigned 32-bit integers in
a
to packed 64-bit integers - Converts packed double-precision (64-bit) floating-point elements in
a
to packed 32-bit integers. - Converts packed double-precision (64-bit) floating-point elements in
a
to packed single-precision (32-bit) floating-point elements - Converts the 4 x 16-bit half-precision float values in the lowest 64-bit of the 128-bit vector
a
into 4 x 32-bit float values stored in a 128-bit wide vector. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed 32-bit integers. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed double-precision (64-bit) floating-point elements. - Converts the 4 x 32-bit float values in the 128-bit vector
a
into 4 x 16-bit half-precision float values stored in the lowest 64-bit of a 128-bit vector. - Returns the lower double-precision (64-bit) floating-point element of
a
. - Converts the lower double-precision (64-bit) floating-point element in a to a 32-bit integer.
- Converts the lower double-precision (64-bit) floating-point element in a to a 64-bit integer.
- Alias for
_mm_cvtsd_si64
- Converts the lower double-precision (64-bit) floating-point element in
b
to a single-precision (32-bit) floating-point element, store the result in the lower element of the return value, and copies the upper element froma
to the upper element the return value. - Returns
a
with its lower element replaced byb
after converting it to anf64
. - Returns a vector whose lowest element is
a
and all higher elements are0
. - Converts a 32 bit integer to a 32 bit float. The result vector is the input vector
a
with the lowest 32 bit float replaced by the converted integer. - Returns
a
with its lower element replaced byb
after converting it to anf64
. - Returns a vector whose lowest element is
a
and all higher elements are0
. - Converts a 64 bit integer to a 32 bit float. The result vector is the input vector
a
with the lowest 32 bit float replaced by the converted integer. - Returns
a
with its lower element replaced byb
after converting it to anf64
. - Returns a vector whose lowest element is
a
and all higher elements are0
. - Returns the lowest element of
a
. - Returns the lowest element of
a
. - Returns the lowest element of
a
. - Extracts the lowest 32 bit float from the input vector.
- Converts the lower single-precision (32-bit) floating-point element in
b
to a double-precision (64-bit) floating-point element, store the result in the lower element of the return value, and copies the upper element froma
to the upper element the return value. - Converts the lowest 32 bit float in the input vector to a 32 bit integer.
- Converts the lowest 32 bit float in the input vector to a 64 bit integer.
- Alias for
_mm_cvttss_si32
. - Converts packed double-precision (64-bit) floating-point elements in
a
to packed 32-bit integers with truncation. - Converts packed single-precision (32-bit) floating-point elements in
a
to packed 32-bit integers with truncation. - Converts the lower double-precision (64-bit) floating-point element in
a
to a 32-bit integer with truncation. - Converts the lower double-precision (64-bit) floating-point element in
a
to a 64-bit integer with truncation. - Alias for
_mm_cvttsd_si64
- Converts the lowest 32 bit float in the input vector to a 32 bit integer with truncation.
- Converts the lowest 32 bit float in the input vector to a 64 bit integer with truncation.
- Divide packed double-precision (64-bit) floating-point elements in
a
by packed elements inb
. - Divides __m128 vectors.
- Returns a new vector with the low element of
a
replaced by the result of diving the lower element ofa
by the lower element ofb
. - Divides the first component of
b
bya
, the other components are copied froma
. - Returns the dot product of two __m128d vectors.
- Returns the dot product of two __m128 vectors.
- Extracts an 8-bit integer from
a
, selected withIMM8
. Returns a 32-bit integer containing the zero-extended integer data. - Returns the
imm8
element ofa
. - Extracts an 32-bit integer from
a
selected withIMM8
- Extracts an 64-bit integer from
a
selected withIMM1
- Extracts a single-precision (32-bit) floating-point element from
a
, selected withIMM8
. The returnedi32
stores the float’s bit-pattern, and may be converted back to a floating point number via casting. - Extracts the bit range specified by
y
from the lower 64 bits ofx
. - Extracts the specified bits from the lower 64 bits of the 128-bit integer vector operand at the index
idx
and of the lengthlen
. - Round the packed double-precision (64-bit) floating-point elements in
a
down to an integer value, and stores the results as packed double-precision floating-point elements. - Round the packed single-precision (32-bit) floating-point elements in
a
down to an integer value, and stores the results as packed single-precision floating-point elements. - Round the lower double-precision (64-bit) floating-point element in
b
down to an integer value, store the result as a double-precision floating-point element in the lower element of the intrinsic result, and copies the upper element froma
to the upper element of the intrinsic result. - Round the lower single-precision (32-bit) floating-point element in
b
down to an integer value, store the result as a single-precision floating-point element in the lower element of the intrinsic result, and copies the upper 3 packed elements froma
to the upper elements of the intrinsic result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and add the intermediate result to packed elements inc
. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and add the intermediate result to packed elements inc
. - Multiplies the lower double-precision (64-bit) floating-point elements in
a
andb
, and add the intermediate result to the lower element inc
. Stores the result in the lower element of the returned value, and copy the upper element froma
to the upper elements of the result. - Multiplies the lower single-precision (32-bit) floating-point elements in
a
andb
, and add the intermediate result to the lower element inc
. Stores the result in the lower element of the returned value, and copy the 3 upper elements froma
to the upper elements of the result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the intermediate result. - Multiplies the lower double-precision (64-bit) floating-point elements in
a
andb
, and subtract the lower element inc
from the intermediate result. Store the result in the lower element of the returned value, and copy the upper element froma
to the upper elements of the result. - Multiplies the lower single-precision (32-bit) floating-point elements in
a
andb
, and subtract the lower element inc
from the intermediate result. Store the result in the lower element of the returned value, and copy the 3 upper elements froma
to the upper elements of the result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and add the negated intermediate result to packed elements inc
. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and add the negated intermediate result to packed elements inc
. - Multiplies the lower double-precision (64-bit) floating-point elements in
a
andb
, and add the negated intermediate result to the lower element inc
. Store the result in the lower element of the returned value, and copy the upper element froma
to the upper elements of the result. - Multiplies the lower single-precision (32-bit) floating-point elements in
a
andb
, and add the negated intermediate result to the lower element inc
. Store the result in the lower element of the returned value, and copy the 3 upper elements froma
to the upper elements of the result. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. - Multiplies packed single-precision (32-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. - Multiplies the lower double-precision (64-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. Store the result in the lower element of the returned value, and copy the upper element froma
to the upper elements of the result. - Multiplies the lower single-precision (32-bit) floating-point elements in
a
andb
, and subtract packed elements inc
from the negated intermediate result. Store the result in the lower element of the returned value, and copy the 3 upper elements froma
to the upper elements of the result. - _mm_
getcsr ⚠Deprecated Gets the unsigned 32-bit value of the MXCSR control and status register. - Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of
[8 x i16]
. - Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of
[4 x i32]
. - Horizontally adds adjacent pairs of double-precision (64-bit) floating-point elements in
a
andb
, and pack the results. - Horizontally adds adjacent pairs of single-precision (32-bit) floating-point elements in
a
andb
, and pack the results. - Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of
[8 x i16]
. Positive sums greater than 7FFFh are saturated to 7FFFh. Negative sums less than 8000h are saturated to 8000h. - Horizontally subtract the adjacent pairs of values contained in 2 packed 128-bit vectors of
[8 x i16]
. - Horizontally subtract the adjacent pairs of values contained in 2 packed 128-bit vectors of
[4 x i32]
. - Horizontally subtract adjacent pairs of double-precision (64-bit) floating-point elements in
a
andb
, and pack the results. - Horizontally adds adjacent pairs of single-precision (32-bit) floating-point elements in
a
andb
, and pack the results. - Horizontally subtract the adjacent pairs of values contained in 2 packed 128-bit vectors of
[8 x i16]
. Positive differences greater than 7FFFh are saturated to 7FFFh. Negative differences less than 8000h are saturated to 8000h. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - Returns a copy of
a
with the 8-bit integer fromi
inserted at a location specified byIMM8
. - Returns a new vector where the
imm8
element ofa
is replaced withi
. - Returns a copy of
a
with the 32-bit integer fromi
inserted at a location specified byIMM8
. - Returns a copy of
a
with the 64-bit integer fromi
inserted at a location specified byIMM1
. - Select a single value in
b
to store at some position ina
, Then zero elements according toIMM8
. - Inserts the
[length:0]
bits ofy
intox
atindex
. - Inserts the
len
least-significant bits from the lower 64 bits of the 128-bit integer vector operandy
into the lower 64 bits of the 128-bit integer vector operandx
at the indexidx
and of the lengthlen
. - Loads 128-bits of integer data from unaligned memory. This intrinsic may perform better than
_mm_loadu_si128
when the data crosses a cache line boundary. - Performs a serializing operation on all load-from-memory instructions that were issued prior to this instruction.
- Loads a double-precision (64-bit) floating-point element from memory into both elements of returned vector.
- Construct a
__m128
by duplicating the value read fromp
into all elements. - Loads 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) from memory into the returned vector.
mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Loads a double-precision (64-bit) floating-point element from memory into both elements of returned vector.
- Loads four
f32
values from aligned memory into a__m128
. If the pointer is not aligned to a 128-bit boundary (16 bytes) a general protection fault will be triggered (fatal program crash). - Alias for
_mm_load1_ps
- Loads a 64-bit double-precision value to the low element of a 128-bit integer vector and clears the upper element.
- Loads 128-bits of integer data from memory into a new vector.
- Construct a
__m128
with the lowest element read fromp
and the other elements set to zero. - Loads a double-precision (64-bit) floating-point element from memory into both elements of return vector.
- Loads a double-precision value into the high-order bits of a 128-bit vector of
[2 x double]
. The low-order bits are copied from the low-order bits of the first operand. - Loads 64-bit integer from memory into first element of returned vector.
- Loads a double-precision value into the low-order bits of a 128-bit vector of
[2 x double]
. The high-order bits are copied from the high-order bits of the first operand. - Loads 2 double-precision (64-bit) floating-point elements from memory into the returned vector in reverse order.
mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Loads four
f32
values from aligned memory into a__m128
in reverse order. - Loads 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) from memory into the returned vector.
mem_addr
does not need to be aligned on any particular boundary. - Loads four
f32
values from memory into a__m128
. There are no restrictions on memory alignment. For aligned memory_mm_load_ps
may be faster. - Loads unaligned 16-bits of integer data from memory into new vector.
- Loads unaligned 32-bits of integer data from memory into new vector.
- Loads unaligned 16-bits of integer data from memory into new vector.
- Loads 128-bits of integer data from memory into a new vector.
- Multiplies and then horizontally add signed 16 bit integers in
a
andb
. - Multiplies corresponding pairs of packed 8-bit unsigned integer values contained in the first source operand and packed 8-bit signed integer values contained in the second source operand, add pairs of contiguous products with signed saturation, and writes the 16-bit sums to the corresponding bits in the destination.
- Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Returns values from
slice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. If mask is set, load the value fromsrc
in that position instead. - Loads packed 32-bit integers from memory pointed by
mem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - Loads packed 64-bit integers from memory pointed by
mem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - Loads packed double-precision (64-bit) floating-point elements from memory into result using
mask
(elements are zeroed out when the high bit of the corresponding element is not set). - Loads packed single-precision (32-bit) floating-point elements from memory into result using
mask
(elements are zeroed out when the high bit of the corresponding element is not set). - Conditionally store 8-bit integer elements from
a
into memory usingmask
. - Stores packed 32-bit integers from
a
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - Stores packed 64-bit integers from
a
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - Stores packed double-precision (64-bit) floating-point elements from
a
into memory usingmask
. - Stores packed single-precision (32-bit) floating-point elements from
a
into memory usingmask
. - Compares packed 8-bit integers in
a
andb
and returns packed maximum values in dst. - Compares packed 16-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed 32-bit integers in
a
andb
, and returns packed maximum values. - Compares packed unsigned 8-bit integers in
a
andb
, and returns the packed maximum values. - Compares packed unsigned 16-bit integers in
a
andb
, and returns packed maximum. - Compares packed unsigned 32-bit integers in
a
andb
, and returns packed maximum values. - Returns a new vector with the maximum values from corresponding elements in
a
andb
. - Compares packed single-precision (32-bit) floating-point elements in
a
andb
, and return the corresponding maximum values. - Returns a new vector with the low element of
a
replaced by the maximum of the lower elements ofa
andb
. - Compares the first single-precision (32-bit) floating-point element of
a
andb
, and return the maximum value in the first element of the return value, the other elements are copied froma
. - Performs a serializing operation on all load-from-memory and store-to-memory instructions that were issued prior to this instruction.
- Compares packed 8-bit integers in
a
andb
and returns packed minimum values in dst. - Compares packed 16-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed 32-bit integers in
a
andb
, and returns packed minimum values. - Compares packed unsigned 8-bit integers in
a
andb
, and returns the packed minimum values. - Compares packed unsigned 16-bit integers in
a
andb
, and returns packed minimum. - Compares packed unsigned 32-bit integers in
a
andb
, and returns packed minimum values. - Returns a new vector with the minimum values from corresponding elements in
a
andb
. - Compares packed single-precision (32-bit) floating-point elements in
a
andb
, and return the corresponding minimum values. - Returns a new vector with the low element of
a
replaced by the minimum of the lower elements ofa
andb
. - Compares the first single-precision (32-bit) floating-point element of
a
andb
, and return the minimum value in the first element of the return value, the other elements are copied froma
. - Finds the minimum unsigned 16-bit element in the 128-bit __m128i vector, returning a vector containing its value in its first position, and its index in its second position; all other elements are set to zero.
- Returns a vector where the low element is extracted from
a
and its upper element is zero. - Constructs a 128-bit floating-point vector of
[2 x double]
. The lower 64 bits are set to the lower 64 bits of the second parameter. The upper 64 bits are set to the upper 64 bits of the first parameter. - Returns a
__m128
with the first component fromb
and the remaining components froma
. - Duplicate the low double-precision (64-bit) floating-point element from
a
. - Duplicate odd-indexed single-precision (32-bit) floating-point elements from
a
. - Combine higher half of
a
andb
. The higher half ofb
occupies the lower half of result. - Duplicate even-indexed single-precision (32-bit) floating-point elements from
a
. - Combine lower half of
a
andb
. The lower half ofb
occupies the higher half of result. - Returns a mask of the most significant bit of each element in
a
. - Returns a mask of the most significant bit of each element in
a
. - Returns a mask of the most significant bit of each element in
a
. - Subtracts 8-bit unsigned integer values and computes the absolute values of the differences to the corresponding bits in the destination. Then sums of the absolute differences are returned according to the bit fields in the immediate operand.
- Multiplies the low 32-bit integers from each packed 64-bit element in
a
andb
, and returns the signed 64-bit result. - Multiplies the low unsigned 32-bit integers from each packed 64-bit element in
a
andb
. - Multiplies packed double-precision (64-bit) floating-point elements in
a
andb
. - Multiplies __m128 vectors.
- Returns a new vector with the low element of
a
replaced by multiplying the low elements ofa
andb
. - Multiplies the first component of
a
andb
, the other components are copied froma
. - Multiplies the packed 16-bit integers in
a
andb
. - Multiplies the packed unsigned 16-bit integers in
a
andb
. - Multiplies packed 16-bit signed integer values, truncate the 32-bit product to the 18 most significant bits by right-shifting, round the truncated value by adding 1, and write bits
[16:1]
to the destination. - Multiplies the packed 16-bit integers in
a
andb
. - Multiplies the packed 32-bit integers in
a
andb
, producing intermediate 64-bit integers, and returns the lowest 32-bit, whatever they might be, reinterpreted as a signed integer. Whilepmulld __m128i::splat(2), __m128i::splat(2)
returns the obvious__m128i::splat(4)
, due to wrapping arithmeticpmulld __m128i::splat(i32::MAX), __m128i::splat(2)
would return a negative number. - Computes the bitwise OR of
a
andb
. - Bitwise OR of packed single-precision (32-bit) floating-point elements.
- Computes the bitwise OR of 128 bits (representing integer data) in
a
andb
. - Converts packed 16-bit integers from
a
andb
to packed 8-bit integers using signed saturation. - Converts packed 32-bit integers from
a
andb
to packed 16-bit integers using signed saturation. - Converts packed 16-bit integers from
a
andb
to packed 8-bit integers using unsigned saturation. - Converts packed 32-bit integers from
a
andb
to packed 16-bit integers using unsigned saturation - Provides a hint to the processor that the code sequence is a spin-wait loop.
- Shuffles double-precision (64-bit) floating-point elements in
a
using the control inimm8
. - Shuffles single-precision (32-bit) floating-point elements in
a
using the control inimm8
. - Shuffles double-precision (64-bit) floating-point elements in
a
using the control inb
. - Shuffles single-precision (32-bit) floating-point elements in
a
using the control inb
. - Fetch the cache line that contains address
p
using the givenSTRATEGY
. - Returns the approximate reciprocal of packed single-precision (32-bit) floating-point elements in
a
. - Returns the approximate reciprocal of the first single-precision (32-bit) floating-point element in
a
, the other elements are unchanged. - Round the packed double-precision (64-bit) floating-point elements in
a
using theROUNDING
parameter, and stores the results as packed double-precision floating-point elements. Rounding is done according to the rounding parameter, which can be one of: - Round the packed single-precision (32-bit) floating-point elements in
a
using theROUNDING
parameter, and stores the results as packed single-precision floating-point elements. Rounding is done according to the rounding parameter, which can be one of: - Round the lower double-precision (64-bit) floating-point element in
b
using theROUNDING
parameter, store the result as a double-precision floating-point element in the lower element of the intrinsic result, and copies the upper element froma
to the upper element of the intrinsic result. Rounding is done according to the rounding parameter, which can be one of: - Round the lower single-precision (32-bit) floating-point element in
b
using theROUNDING
parameter, store the result as a single-precision floating-point element in the lower element of the intrinsic result, and copies the upper 3 packed elements froma
to the upper elements of the intrinsic result. Rounding is done according to the rounding parameter, which can be one of: - Returns the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in
a
. - Returns the approximate reciprocal square root of the first single-precision (32-bit) floating-point element in
a
, the other elements are unchanged. - Sum the absolute differences of packed unsigned 8-bit integers.
- Broadcasts 8-bit integer
a
to all elements. - Broadcasts 16-bit integer
a
to all elements. - Broadcasts 32-bit integer
a
to all elements. - Broadcasts 64-bit integer
a
to all elements. - Broadcasts double-precision (64-bit) floating-point value a to all elements of the return value.
- Construct a
__m128
with all element set toa
. - Sets packed 8-bit integers with the supplied values.
- Sets packed 16-bit integers with the supplied values.
- Sets packed 32-bit integers with the supplied values.
- Sets packed 64-bit integers with the supplied values, from highest to lowest.
- Sets packed double-precision (64-bit) floating-point elements in the return value with the supplied values.
- Broadcasts double-precision (64-bit) floating-point value a to all elements of the return value.
- Construct a
__m128
from four floating point values highest to lowest. - Alias for
_mm_set1_ps
- Copies double-precision (64-bit) floating-point element
a
to the lower element of the packed 64-bit return value. - Construct a
__m128
with the lowest element set toa
and the rest set to zero. - _mm_
setcsr ⚠Deprecated Sets the MXCSR register with the 32-bit unsigned integer value. - Sets packed 8-bit integers with the supplied values in reverse order.
- Sets packed 16-bit integers with the supplied values in reverse order.
- Sets packed 32-bit integers with the supplied values in reverse order.
- Sets packed double-precision (64-bit) floating-point elements in the return value with the supplied values in reverse order.
- Construct a
__m128
from four floating point values lowest to highest. - Returns packed double-precision (64-bit) floating-point elements with all zeros.
- Construct a
__m128
with all elements initialized to zero. - Returns a vector with all elements set to zero.
- Performs a serializing operation on all non-temporal (“streaming”) store instructions that were issued by the current thread prior to this instruction.
- Performs an intermediate calculation for the next four SHA1 message values (unsigned 32-bit integers) using previous message values from
a
andb
, and returning the result. - Performs the final calculation for the next four SHA1 message values (unsigned 32-bit integers) using the intermediate result in
a
and the previous message values inb
, and returns the result. - Calculate SHA1 state variable E after four rounds of operation from the current SHA1 state variable
a
, add that value to the scheduled values (unsigned 32-bit integers) inb
, and returns the result. - Performs four rounds of SHA1 operation using an initial SHA1 state (A,B,C,D) from
a
and some pre-computed sum of the next 4 round message values (unsigned 32-bit integers), and state variable E fromb
, and return the updated SHA1 state (A,B,C,D).FUNC
contains the logic functions and round constants. - Performs an intermediate calculation for the next four SHA256 message values (unsigned 32-bit integers) using previous message values from
a
andb
, and return the result. - Performs the final calculation for the next four SHA256 message values (unsigned 32-bit integers) using previous message values from
a
andb
, and return the result. - Performs 2 rounds of SHA256 operation using an initial SHA256 state (C,D,G,H) from
a
, an initial SHA256 state (A,B,E,F) fromb
, and a pre-computed sum of the next 2 round message values (unsigned 32-bit integers) and the corresponding round constants fromk
, and store the updated SHA256 state (A,B,E,F) in dst. - Shuffles bytes from
a
according to the content ofb
. - Shuffles 32-bit integers in
a
using the control inIMM8
. - Constructs a 128-bit floating-point vector of
[2 x double]
from two 128-bit vector parameters of[2 x double]
, using the immediate-value parameter as a specifier. - Shuffles packed single-precision (32-bit) floating-point elements in
a
andb
usingMASK
. - Shuffles 16-bit integers in the high 64 bits of
a
using the control inIMM8
. - Shuffles 16-bit integers in the low 64 bits of
a
using the control inIMM8
. - Negates packed 8-bit integers in
a
when the corresponding signed 8-bit integer inb
is negative, and returns the result. Elements in result are zeroed out when the corresponding element inb
is zero. - Negates packed 16-bit integers in
a
when the corresponding signed 16-bit integer inb
is negative, and returns the results. Elements in result are zeroed out when the corresponding element inb
is zero. - Negates packed 32-bit integers in
a
when the corresponding signed 32-bit integer inb
is negative, and returns the results. Element in result are zeroed out when the corresponding element inb
is zero. - Shifts packed 16-bit integers in
a
left bycount
while shifting in zeros. - Shifts packed 32-bit integers in
a
left bycount
while shifting in zeros. - Shifts packed 64-bit integers in
a
left bycount
while shifting in zeros. - Shifts packed 16-bit integers in
a
left byIMM8
while shifting in zeros. - Shifts packed 32-bit integers in
a
left byIMM8
while shifting in zeros. - Shifts packed 64-bit integers in
a
left byIMM8
while shifting in zeros. - Shifts
a
left byIMM8
bytes while shifting in zeros. - Shifts packed 32-bit integers in
a
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - Shifts packed 64-bit integers in
a
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - Returns a new vector with the square root of each of the values in
a
. - Returns the square root of packed single-precision (32-bit) floating-point elements in
a
. - Returns a new vector with the low element of
a
replaced by the square root of the lower elementb
. - Returns the square root of the first single-precision (32-bit) floating-point element in
a
, the other elements are unchanged. - Shifts packed 16-bit integers in
a
right bycount
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right bycount
while shifting in sign bits. - Shifts packed 16-bit integers in
a
right byIMM8
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right byIMM8
while shifting in sign bits. - Shifts packed 32-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in sign bits. - Shifts packed 16-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 32-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 64-bit integers in
a
right bycount
while shifting in zeros. - Shifts packed 16-bit integers in
a
right byIMM8
while shifting in zeros. - Shifts packed 32-bit integers in
a
right byIMM8
while shifting in zeros. - Shifts packed 64-bit integers in
a
right byIMM8
while shifting in zeros. - Shifts
a
right byIMM8
bytes while shifting in zeros. - Shifts packed 32-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in zeros, - Shifts packed 64-bit integers in
a
right by the amount specified by the corresponding element incount
while shifting in zeros, - Stores the lower double-precision (64-bit) floating-point element from
a
into 2 contiguous elements in memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Stores the lowest 32 bit float of
a
repeated four times into aligned memory. - Stores 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) from
a
into memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Stores the lower double-precision (64-bit) floating-point element from
a
into 2 contiguous elements in memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Stores four 32-bit floats into aligned memory.
- Alias for
_mm_store1_ps
- Stores the lower 64 bits of a 128-bit vector of
[2 x double]
to a memory location. - Stores 128-bits of integer data from
a
into memory. - Stores the lowest 32 bit float of
a
into memory. - Stores the upper 64 bits of a 128-bit vector of
[2 x double]
to a memory location. - Stores the lower 64-bit integer
a
to a memory location. - Stores the lower 64 bits of a 128-bit vector of
[2 x double]
to a memory location. - Stores 2 double-precision (64-bit) floating-point elements from
a
into memory in reverse order.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - Stores four 32-bit floats into aligned memory in reverse order.
- Stores 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) from
a
into memory.mem_addr
does not need to be aligned on any particular boundary. - Stores four 32-bit floats into memory. There are no restrictions on memory alignment. For aligned memory
_mm_store_ps
may be faster. - Store 16-bit integer from the first element of a into memory.
- Store 32-bit integer from the first element of a into memory.
- Store 64-bit integer from the first element of a into memory.
- Stores 128-bits of integer data from
a
into memory. - Load 128-bits of integer data from memory into dst. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon)
- Stores a 128-bit floating point vector of
[2 x double]
to a 128-bit aligned memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon). - Stores
a
into the memory atmem_addr
using a non-temporal memory hint. - Non-temporal store of
a.0
intop
. - Stores a 32-bit integer value in the specified memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon).
- Stores a 64-bit integer value in the specified memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon).
- Stores a 128-bit integer vector to a 128-bit aligned memory location. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon).
- Non-temporal store of
a.0
intop
. - Subtracts packed 8-bit integers in
b
from packed 8-bit integers ina
. - Subtracts packed 16-bit integers in
b
from packed 16-bit integers ina
. - Subtract packed 32-bit integers in
b
from packed 32-bit integers ina
. - Subtract packed 64-bit integers in
b
from packed 64-bit integers ina
. - Subtract packed double-precision (64-bit) floating-point elements in
b
froma
. - Subtracts __m128 vectors.
- Returns a new vector with the low element of
a
replaced by subtracting the low element byb
from the low element ofa
. - Subtracts the first component of
b
froma
, the other components are copied froma
. - Subtract packed 8-bit integers in
b
from packed 8-bit integers ina
using saturation. - Subtract packed 16-bit integers in
b
from packed 16-bit integers ina
using saturation. - Subtract packed unsigned 8-bit integers in
b
from packed unsigned 8-bit integers ina
using saturation. - Subtract packed unsigned 16-bit integers in
b
from packed unsigned 16-bit integers ina
using saturation. - Tests whether the specified bits in
a
128-bit integer vector are all ones. - Tests whether the specified bits in a 128-bit integer vector are all zeros.
- Tests whether the specified bits in a 128-bit integer vector are neither all zeros nor all ones.
- Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theCF
value. - Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theCF
value. - Tests whether the specified bits in a 128-bit integer vector are all ones.
- Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return 1 if both theZF
andCF
values are zero, otherwise return 0. - Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return 1 if both theZF
andCF
values are zero, otherwise return 0. - Tests whether the specified bits in a 128-bit integer vector are neither all zeros nor all ones.
- Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 64-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theZF
value. - Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) in
a
andb
, producing an intermediate 128-bit value, and setZF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setZF
to 0. Compute the bitwise NOT ofa
and then AND withb
, producing an intermediate value, and setCF
to 1 if the sign bit of each 32-bit element in the intermediate value is zero, otherwise setCF
to 0. Return theZF
value. - Tests whether the specified bits in a 128-bit integer vector are all zeros.
- Counts the number of trailing least significant zero bits.
- Counts the number of trailing least significant zero bits.
- Compares the lower element of
a
andb
for equality. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if they are equal, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Compares the lower element of
a
andb
for greater-than-or-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is greater than or equal to the one fromb
, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Compares the lower element of
a
andb
for greater-than. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is greater than the one fromb
, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Compares the lower element of
a
andb
for less-than-or-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is less than or equal to the one fromb
, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Compares the lower element of
a
andb
for less-than. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if the value froma
is less than the one fromb
, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Compares the lower element of
a
andb
for not-equal. - Compares two 32-bit floats from the low-order bits of
a
andb
. Returns1
if they are not equal, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - Returns vector of type __m128d with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent to
mem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Returns vector of type __m128 with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent to
mem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Returns vector of type __m128i with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent to
mem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - Unpacks and interleave 8-bit integers from the high half of
a
andb
. - Unpacks and interleave 16-bit integers from the high half of
a
andb
. - Unpacks and interleave 32-bit integers from the high half of
a
andb
. - Unpacks and interleave 64-bit integers from the high half of
a
andb
. - The resulting
__m128d
element is composed by the low-order values of the two__m128d
interleaved input elements, i.e.: - Unpacks and interleave single-precision (32-bit) floating-point elements from the higher half of
a
andb
. - Unpacks and interleave 8-bit integers from the low half of
a
andb
. - Unpacks and interleave 16-bit integers from the low half of
a
andb
. - Unpacks and interleave 32-bit integers from the low half of
a
andb
. - Unpacks and interleave 64-bit integers from the low half of
a
andb
. - The resulting
__m128d
element is composed by the high-order values of the two__m128d
interleaved input elements, i.e.: - Unpacks and interleave single-precision (32-bit) floating-point elements from the lower half of
a
andb
. - Computes the bitwise XOR of
a
andb
. - Bitwise exclusive OR of packed single-precision (32-bit) floating-point elements.
- Computes the bitwise XOR of 128 bits (representing integer data) in
a
andb
. - Unsigned multiply without affecting flags.
- Unsigned multiply without affecting flags.
- Scatter contiguous low order bits of
a
to the result at the positions specified by themask
. - Scatter contiguous low order bits of
a
to the result at the positions specified by themask
. - Gathers the bits of
x
specified by themask
into the contiguous low order bit positions of the result. - Gathers the bits of
x
specified by themask
into the contiguous low order bit positions of the result. - Counts the bits that are set.
- Counts the bits that are set.
- Read a hardware generated 16-bit random value and store the result in val. Returns 1 if a random value was generated, and 0 otherwise.
- Read a hardware generated 32-bit random value and store the result in val. Returns 1 if a random value was generated, and 0 otherwise.
- Read a hardware generated 64-bit random value and store the result in val. Returns 1 if a random value was generated, and 0 otherwise.
- Read a 16-bit NIST SP800-90B and SP800-90C compliant random value and store in val. Return 1 if a random value was generated, and 0 otherwise.
- Read a 32-bit NIST SP800-90B and SP800-90C compliant random value and store in val. Return 1 if a random value was generated, and 0 otherwise.
- Read a 64-bit NIST SP800-90B and SP800-90C compliant random value and store in val. Return 1 if a random value was generated, and 0 otherwise.
- Reads the current value of the processor’s time-stamp counter.
- Adds unsigned 32-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry or overflow flag), and store the unsigned 32-bit result inout
, and the carry-out is returned (carry or overflow flag). - Adds unsigned 64-bit integers
a
andb
with unsigned 8-bit carry-inc_in
. (carry or overflow flag), and store the unsigned 64-bit result inout
, and the carry-out is returned (carry or overflow flag). - Clears all bits below the least significant zero of
x
and sets all other bits. - Clears all bits below the least significant zero of
x
and sets all other bits. - Counts the number of trailing least significant zero bits.
- Counts the number of trailing least significant zero bits.
- Counts the number of trailing least significant zero bits.
- Sets all bits below the least significant one of
x
and clears all other bits. - Sets all bits below the least significant one of
x
and clears all other bits. - Reads the contents of the extended control register
XCR
specified inxcr_no
. - Performs a full or partial restore of the enabled processor states using the state information stored in memory at
mem_addr
. - Performs a full or partial restore of the enabled processor states using the state information stored in memory at
mem_addr
. - Performs a full or partial restore of the enabled processor states using the state information stored in memory at
mem_addr
. - Performs a full or partial restore of the enabled processor states using the state information stored in memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
. - Performs a full or partial save of the enabled processor states to memory at
mem_addr
- Performs a full or partial save of the enabled processor states to memory at
mem_addr
- Copies 64-bits from
val
to the extended control register (XCR
) specified bya
. - Compares and exchange 16 bytes (128 bits) of data atomically.
- _MM_
SHUFFLE Experimental A utility function for creating masks to use with Intel shuffle and permute intrinsics. - _cvtmask8_
u32 ⚠Experimental Convert 8-bit mask a to a 32-bit integer value and store the result in dst. - _cvtmask16_
u32 ⚠Experimental Convert 16-bit mask a into an integer value, and store the result in dst. - _cvtmask32_
u32 ⚠Experimental Convert 32-bit mask a into an integer value, and store the result in dst. - _cvtmask64_
u64 ⚠Experimental Convert 64-bit mask a into an integer value, and store the result in dst. - _cvtu32_
mask8 ⚠Experimental Convert 32-bit integer value a to an 8-bit mask and store the result in dst. - _cvtu32_
mask16 ⚠Experimental Convert 32-bit integer value a to an 16-bit mask and store the result in dst. - _cvtu32_
mask32 ⚠Experimental Convert integer value a into an 32-bit mask, and store the result in k. - _cvtu64_
mask64 ⚠Experimental Convert integer value a into an 64-bit mask, and store the result in k. - _kadd_
mask8 ⚠Experimental Add 8-bit masks a and b, and store the result in dst. - _kadd_
mask16 ⚠Experimental Add 16-bit masks a and b, and store the result in dst. - _kadd_
mask32 ⚠Experimental Add 32-bit masks in a and b, and store the result in k. - _kadd_
mask64 ⚠Experimental Add 64-bit masks in a and b, and store the result in k. - _kand_
mask8 ⚠Experimental Bitwise AND of 8-bit masks a and b, and store the result in dst. - _kand_
mask16 ⚠Experimental Compute the bitwise AND of 16-bit masks a and b, and store the result in k. - _kand_
mask32 ⚠Experimental Compute the bitwise AND of 32-bit masks a and b, and store the result in k. - _kand_
mask64 ⚠Experimental Compute the bitwise AND of 64-bit masks a and b, and store the result in k. - _kandn_
mask8 ⚠Experimental Bitwise AND NOT of 8-bit masks a and b, and store the result in dst. - _kandn_
mask16 ⚠Experimental Compute the bitwise NOT of 16-bit masks a and then AND with b, and store the result in k. - _kandn_
mask32 ⚠Experimental Compute the bitwise NOT of 32-bit masks a and then AND with b, and store the result in k. - _kandn_
mask64 ⚠Experimental Compute the bitwise NOT of 64-bit masks a and then AND with b, and store the result in k. - _knot_
mask8 ⚠Experimental Bitwise NOT of 8-bit mask a, and store the result in dst. - _knot_
mask16 ⚠Experimental Compute the bitwise NOT of 16-bit mask a, and store the result in k. - _knot_
mask32 ⚠Experimental Compute the bitwise NOT of 32-bit mask a, and store the result in k. - _knot_
mask64 ⚠Experimental Compute the bitwise NOT of 64-bit mask a, and store the result in k. - _kor_
mask8 ⚠Experimental Bitwise OR of 8-bit masks a and b, and store the result in dst. - _kor_
mask16 ⚠Experimental Compute the bitwise OR of 16-bit masks a and b, and store the result in k. - _kor_
mask32 ⚠Experimental Compute the bitwise OR of 32-bit masks a and b, and store the result in k. - _kor_
mask64 ⚠Experimental Compute the bitwise OR of 64-bit masks a and b, and store the result in k. - _kortest_
mask8_ ⚠u8 Experimental Compute the bitwise OR of 8-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. If the result is all ones, store 1 in all_ones, otherwise store 0 in all_ones. - _kortest_
mask16_ ⚠u8 Experimental Compute the bitwise OR of 16-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. If the result is all ones, store 1 in all_ones, otherwise store 0 in all_ones. - _kortest_
mask32_ ⚠u8 Experimental Compute the bitwise OR of 32-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. If the result is all ones, store 1 in all_ones, otherwise store 0 in all_ones. - _kortest_
mask64_ ⚠u8 Experimental Compute the bitwise OR of 64-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. If the result is all ones, store 1 in all_ones, otherwise store 0 in all_ones. - _kortestc_
mask8_ ⚠u8 Experimental Compute the bitwise OR of 8-bit masks a and b. If the result is all ones, store 1 in dst, otherwise store 0 in dst. - _kortestc_
mask16_ ⚠u8 Experimental Compute the bitwise OR of 16-bit masks a and b. If the result is all ones, store 1 in dst, otherwise store 0 in dst. - _kortestc_
mask32_ ⚠u8 Experimental Compute the bitwise OR of 32-bit masks a and b. If the result is all ones, store 1 in dst, otherwise store 0 in dst. - _kortestc_
mask64_ ⚠u8 Experimental Compute the bitwise OR of 64-bit masks a and b. If the result is all ones, store 1 in dst, otherwise store 0 in dst. - _kortestz_
mask8_ ⚠u8 Experimental Compute the bitwise OR of 8-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _kortestz_
mask16_ ⚠u8 Experimental Compute the bitwise OR of 16-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _kortestz_
mask32_ ⚠u8 Experimental Compute the bitwise OR of 32-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _kortestz_
mask64_ ⚠u8 Experimental Compute the bitwise OR of 64-bit masks a and b. If the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _kshiftli_
mask8 ⚠Experimental Shift 8-bit mask a left by count bits while shifting in zeros, and store the result in dst. - _kshiftli_
mask16 ⚠Experimental Shift 16-bit mask a left by count bits while shifting in zeros, and store the result in dst. - _kshiftli_
mask32 ⚠Experimental Shift the bits of 32-bit mask a left by count while shifting in zeros, and store the least significant 32 bits of the result in k. - _kshiftli_
mask64 ⚠Experimental Shift the bits of 64-bit mask a left by count while shifting in zeros, and store the least significant 32 bits of the result in k. - _kshiftri_
mask8 ⚠Experimental Shift 8-bit mask a right by count bits while shifting in zeros, and store the result in dst. - _kshiftri_
mask16 ⚠Experimental Shift 16-bit mask a right by count bits while shifting in zeros, and store the result in dst. - _kshiftri_
mask32 ⚠Experimental Shift the bits of 32-bit mask a right by count while shifting in zeros, and store the least significant 32 bits of the result in k. - _kshiftri_
mask64 ⚠Experimental Shift the bits of 64-bit mask a right by count while shifting in zeros, and store the least significant 32 bits of the result in k. - _ktest_
mask8_ ⚠u8 Experimental Compute the bitwise AND of 8-bit masks a and b, and if the result is all zeros, store 1 in dst, otherwise store 0 in dst. Compute the bitwise NOT of a and then AND with b, if the result is all zeros, store 1 in and_not, otherwise store 0 in and_not. - _ktest_
mask16_ ⚠u8 Experimental Compute the bitwise AND of 16-bit masks a and b, and if the result is all zeros, store 1 in dst, otherwise store 0 in dst. Compute the bitwise NOT of a and then AND with b, if the result is all zeros, store 1 in and_not, otherwise store 0 in and_not. - _ktest_
mask32_ ⚠u8 Experimental Compute the bitwise AND of 32-bit masks a and b, and if the result is all zeros, store 1 in dst, otherwise store 0 in dst. Compute the bitwise NOT of a and then AND with b, if the result is all zeros, store 1 in and_not, otherwise store 0 in and_not. - _ktest_
mask64_ ⚠u8 Experimental Compute the bitwise AND of 64-bit masks a and b, and if the result is all zeros, store 1 in dst, otherwise store 0 in dst. Compute the bitwise NOT of a and then AND with b, if the result is all zeros, store 1 in and_not, otherwise store 0 in and_not. - _ktestc_
mask8_ ⚠u8 Experimental Compute the bitwise NOT of 8-bit mask a and then AND with 8-bit mask b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestc_
mask16_ ⚠u8 Experimental Compute the bitwise NOT of 16-bit mask a and then AND with 16-bit mask b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestc_
mask32_ ⚠u8 Experimental Compute the bitwise NOT of 32-bit mask a and then AND with 16-bit mask b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestc_
mask64_ ⚠u8 Experimental Compute the bitwise NOT of 64-bit mask a and then AND with 8-bit mask b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestz_
mask8_ ⚠u8 Experimental Compute the bitwise AND of 8-bit masks a and b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestz_
mask16_ ⚠u8 Experimental Compute the bitwise AND of 16-bit masks a and b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestz_
mask32_ ⚠u8 Experimental Compute the bitwise AND of 32-bit masks a and b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _ktestz_
mask64_ ⚠u8 Experimental Compute the bitwise AND of 64-bit masks a and b, if the result is all zeros, store 1 in dst, otherwise store 0 in dst. - _kxnor_
mask8 ⚠Experimental Bitwise XNOR of 8-bit masks a and b, and store the result in dst. - _kxnor_
mask16 ⚠Experimental Compute the bitwise XNOR of 16-bit masks a and b, and store the result in k. - _kxnor_
mask32 ⚠Experimental Compute the bitwise XNOR of 32-bit masks a and b, and store the result in k. - _kxnor_
mask64 ⚠Experimental Compute the bitwise XNOR of 64-bit masks a and b, and store the result in k. - _kxor_
mask8 ⚠Experimental Bitwise XOR of 8-bit masks a and b, and store the result in dst. - _kxor_
mask16 ⚠Experimental Compute the bitwise XOR of 16-bit masks a and b, and store the result in k. - _kxor_
mask32 ⚠Experimental Compute the bitwise XOR of 32-bit masks a and b, and store the result in k. - _kxor_
mask64 ⚠Experimental Compute the bitwise XOR of 64-bit masks a and b, and store the result in k. - _load_
mask8 ⚠Experimental Load 8-bit mask from memory - _load_
mask16 ⚠Experimental Load 16-bit mask from memory - _load_
mask32 ⚠Experimental Load 32-bit mask from memory into k. - _load_
mask64 ⚠Experimental Load 64-bit mask from memory into k. - _mm256_
abs_ ⚠epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst. - _mm256_
abs_ ⚠ph Experimental Finds the absolute value of each packed half-precision (16-bit) floating-point element in v2, storing the result in dst. - _mm256_
add_ ⚠ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm256_
aesdec_ ⚠epi128 Experimental Performs one round of an AES decryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm256_
aesdeclast_ ⚠epi128 Experimental Performs the last round of an AES decryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm256_
aesenc_ ⚠epi128 Experimental Performs one round of an AES encryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm256_
aesenclast_ ⚠epi128 Experimental Performs the last round of an AES encryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm256_
alignr_ ⚠epi32 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 32 bytes (8 elements) in dst. - _mm256_
alignr_ ⚠epi64 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 32 bytes (4 elements) in dst. - _mm256_
bcstnebf16_ ⚠ps Experimental Convert scalar BF16 (16-bit) floating point element stored at memory locations starting at location a to single precision (32-bit) floating-point, broadcast it to packed single precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
bcstnesh_ ⚠ps Experimental Convert scalar half-precision (16-bit) floating-point element stored at memory locations starting at location a to a single-precision (32-bit) floating-point, broadcast it to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
bitshuffle_ ⚠epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm256_
broadcast_ ⚠f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst. - _mm256_
broadcast_ ⚠f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst. - _mm256_
broadcast_ ⚠f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst. - _mm256_
broadcast_ ⚠i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst. - _mm256_
broadcast_ ⚠i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst. - _mm256_
broadcast_ ⚠i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst. - _mm256_
broadcastmb_ ⚠epi64 Experimental Broadcast the low 8-bits from input mask k to all 64-bit elements of dst. - _mm256_
broadcastmw_ ⚠epi32 Experimental Broadcast the low 16-bits from input mask k to all 32-bit elements of dst. - _mm256_
castpd_ ⚠ph Experimental Cast vector of type__m256d
to type__m256h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castph128_ ⚠ph256 Experimental Cast vector of type__m128h
to type__m256h
. The upper 8 elements of the result are undefined. In practice, the upper elements are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm256_
castph256_ ⚠ph128 Experimental Cast vector of type__m256h
to type__m128h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castph_ ⚠pd Experimental Cast vector of type__m256h
to type__m256d
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castph_ ⚠ps Experimental Cast vector of type__m256h
to type__m256
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castph_ ⚠si256 Experimental Cast vector of type__m256h
to type__m256i
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castps_ ⚠ph Experimental Cast vector of type__m256
to type__m256h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
castsi256_ ⚠ph Experimental Cast vector of type__m256i
to type__m256h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm256_
clmulepi64_ ⚠epi128 Experimental Performs a carry-less multiplication of two 64-bit polynomials over the finite field GF(2) - in each of the 2 128-bit lanes. - _mm256_
cmp_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmp_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpeq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpge_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmpgt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm256_
cmple_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmple_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmplt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmpneq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm256_
cmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
conflict_ ⚠epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
conflict_ ⚠epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
conj_ ⚠pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
cvtepi16_ ⚠epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi16_ ⚠ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepi32_ ⚠epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi32_ ⚠epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi32_ ⚠ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepi64_ ⚠epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi64_ ⚠epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi64_ ⚠epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst. - _mm256_
cvtepi64_ ⚠pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepi64_ ⚠ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 64 bits of dst are zeroed out. - _mm256_
cvtepi64_ ⚠ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepu16_ ⚠ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepu32_ ⚠pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepu32_ ⚠ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepu64_ ⚠pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm256_
cvtepu64_ ⚠ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 64 bits of dst are zeroed out. - _mm256_
cvtepu64_ ⚠ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtne2ps_ ⚠pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two 256-bit vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in a 256-bit wide vector. Intel’s documentation - _mm256_
cvtneebf16_ ⚠ps Experimental Convert packed BF16 (16-bit) floating-point even-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtneeph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point even-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtneobf16_ ⚠ps Experimental Convert packed BF16 (16-bit) floating-point odd-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtneoph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point odd-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtneps_ ⚠avx_ pbh Experimental Convert packed single precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst. - _mm256_
cvtneps_ ⚠pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst. Intel’s documentation - _mm256_
cvtpbh_ ⚠ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm256_
cvtpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm256_
cvtpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm256_
cvtpd_ ⚠ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 64 bits of dst are zeroed out. - _mm256_
cvtph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst. - _mm256_
cvtph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm256_
cvtph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst. - _mm256_
cvtph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst. - _mm256_
cvtph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst. - _mm256_
cvtph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst. - _mm256_
cvtph_ ⚠pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm256_
cvtps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm256_
cvtps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm256_
cvtps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm256_
cvtsepi16_ ⚠epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsepi32_ ⚠epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsepi32_ ⚠epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsepi64_ ⚠epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsepi64_ ⚠epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsepi64_ ⚠epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst. - _mm256_
cvtsh_ ⚠h Experimental Copy the lower half-precision (16-bit) floating-point element froma
todst
. - _mm256_
cvttpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm256_
cvttpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm256_
cvttpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm256_
cvttph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst. - _mm256_
cvttph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm256_
cvttph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst. - _mm256_
cvttph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst. - _mm256_
cvttph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst. - _mm256_
cvttph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst. - _mm256_
cvttps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm256_
cvttps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm256_
cvttps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm256_
cvtusepi16_ ⚠epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtusepi32_ ⚠epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtusepi32_ ⚠epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtusepi64_ ⚠epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtusepi64_ ⚠epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtusepi64_ ⚠epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst. - _mm256_
cvtxph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm256_
cvtxps_ ⚠ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm256_
dbsad_ ⚠epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst. Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm256_
div_ ⚠ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst. - _mm256_
dpbf16_ ⚠ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst. Intel’s documentation - _mm256_
dpbssd_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpbssds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpbsud_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpbsuds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpbusd_ ⚠avx_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpbusd_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpbusds_ ⚠avx_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpbusds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpbuud_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpbuuds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpwssd_ ⚠avx_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpwssd_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpwssds_ ⚠avx_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpwssds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpwsud_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpwsuds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpwusd_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding signed 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpwusds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding signed 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
dpwuud_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm256_
dpwuuds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm256_
extractf32x4_ ⚠ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the result in dst. - _mm256_
extractf64x2_ ⚠pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst. - _mm256_
extracti32x4_ ⚠epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM1, and store the result in dst. - _mm256_
extracti64x2_ ⚠epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst. - _mm256_
fcmadd_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
fcmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
fixupimm_ ⚠pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm256_
fixupimm_ ⚠ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm256_
fmadd_ ⚠pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
fmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm256_
fmaddsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm256_
fmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. - _mm256_
fmsubadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst. - _mm256_
fmul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
fnmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst. - _mm256_
fnmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm256_
fpclass_ ⚠pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm256_
fpclass_ ⚠ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm256_
fpclass_ ⚠ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm256_
getexp_ ⚠pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
getexp_ ⚠ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm256_
getexp_ ⚠ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
getmant_ ⚠pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
getmant_ ⚠ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm256_
getmant_ ⚠ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign. The mantissa is normalized to the interval specified by interv, which can take the following values: _MM_MANT_NORM_1_2 // interval [1, 2) _MM_MANT_NORM_p5_2 // interval [0.5, 2) _MM_MANT_NORM_p5_1 // interval [0.5, 1) _MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5) The sign is determined by sc which can take the following values: _MM_MANT_SIGN_src // sign = sign(src) _MM_MANT_SIGN_zero // sign = 0 _MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
gf2p8affine_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
gf2p8affineinv_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
gf2p8mul_ ⚠epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm256_
i32scatter_ ⚠epi32 Experimental Stores 8 32-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm256_
i32scatter_ ⚠epi64 Experimental Scatter 64-bit integers from a into memory using 32-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm256_
i32scatter_ ⚠pd Experimental Stores 4 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm256_
i32scatter_ ⚠ps Experimental Stores 8 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm256_
i64scatter_ ⚠epi32 Experimental Stores 4 32-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm256_
i64scatter_ ⚠epi64 Experimental Stores 4 64-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm256_
i64scatter_ ⚠pd Experimental Stores 4 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm256_
i64scatter_ ⚠ps Experimental Stores 4 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm256_
insertf32x4 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into dst at the location specified by imm8. - _mm256_
insertf64x2 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into dst at the location specified by IMM8. - _mm256_
inserti32x4 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 4 packed 32-bit integers) from b into dst at the location specified by imm8. - _mm256_
inserti64x2 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 2 packed 64-bit integers) from b into dst at the location specified by IMM8. - _mm256_
load_ ⚠epi32 Experimental Load 256-bits (composed of 8 packed 32-bit integers) from memory into dst. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
load_ ⚠epi64 Experimental Load 256-bits (composed of 4 packed 64-bit integers) from memory into dst. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
load_ ⚠ph Experimental Load 256-bits (composed of 16 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address must be aligned to 32 bytes or a general-protection exception may be generated. - _mm256_
loadu_ ⚠epi8 Experimental Load 256-bits (composed of 32 packed 8-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm256_
loadu_ ⚠epi16 Experimental Load 256-bits (composed of 16 packed 16-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm256_
loadu_ ⚠epi32 Experimental Load 256-bits (composed of 8 packed 32-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm256_
loadu_ ⚠epi64 Experimental Load 256-bits (composed of 4 packed 64-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm256_
loadu_ ⚠ph Experimental Load 256-bits (composed of 16 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address does not need to be aligned to any particular boundary. - _mm256_
lzcnt_ ⚠epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst. - _mm256_
lzcnt_ ⚠epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst. - _mm256_
madd52hi_ ⚠avx_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm256_
madd52hi_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm256_
madd52lo_ ⚠avx_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm256_
madd52lo_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm256_
mask2_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask2_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm256_
mask2_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm256_
mask2_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm256_
mask2_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set) - _mm256_
mask2_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
mask3_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
mask3_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmaddsub_ pd Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm256_
mask3_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm256_
mask_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠abs_ epi32 Experimental Compute the absolute value of packed signed 32-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 32 bytes (8 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 32 bytes (4 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠and_ epi32 Experimental Performs element-by-element bitwise AND between packed 32-bit integer elements of a and b, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠bitshuffle_ epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm256_
mask_ ⚠blend_ epi8 Experimental Blend packed 8-bit integers from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ epi16 Experimental Blend packed 16-bit integers from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ epi32 Experimental Blend packed 32-bit integers from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ epi64 Experimental Blend packed 64-bit integers from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ pd Experimental Blend packed double-precision (64-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ ph Experimental Blend packed half-precision (16-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠blend_ ps Experimental Blend packed single-precision (32-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm256_
mask_ ⚠broadcast_ f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠broadcast_ f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcast_ f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠broadcast_ i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcast_ i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcastsd_ pd Experimental Broadcast the low double-precision (64-bit) floating-point element from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmp_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpeq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpge_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpgt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmple_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmplt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmpneq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
mask_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm256_
mask_ ⚠compressstoreu_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠compressstoreu_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠compressstoreu_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠compressstoreu_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠compressstoreu_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠compressstoreu_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
mask_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
mask_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
mask_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of: (_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
mask_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi16_ storeu_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi32_ storeu_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepi32_ storeu_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
mask_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtepi64_ storeu_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepi64_ storeu_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepi64_ storeu_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in the low 8 bytes of a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
mask_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements and store the results in single vector dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm256_
mask_ ⚠cvtneps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm256_
mask_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
mask_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
mask_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi16_ storeu_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi32_ storeu_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtsepi32_ storeu_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtsepi64_ storeu_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtsepi64_ storeu_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtsepi64_ storeu_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi16_ storeu_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi32_ storeu_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtusepi32_ storeu_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtusepi64_ storeu_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtusepi64_ storeu_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtusepi64_ storeu_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed 32-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm256_
mask_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm256_
mask_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm256_
mask_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm256_
mask_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠extractf32x4_ ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠extractf64x2_ pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠extracti32x4_ epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM1, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠extracti64x2_ epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
mask_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
mask_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm256_
mask_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm256_
mask_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
mask_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
mask_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠fpclass_ pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm256_
mask_ ⚠fpclass_ ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm256_
mask_ ⚠fpclass_ ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm256_
mask_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
mask_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm256_
mask_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
mask_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
mask_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm256_
mask_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
mask_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
mask_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
mask_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm256_
mask_ ⚠i32scatter_ epi32 Experimental Stores 8 32-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i32scatter_ epi64 Experimental Stores 4 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i32scatter_ pd Experimental Stores 4 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i32scatter_ ps Experimental Stores 8 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i64scatter_ epi32 Experimental Stores 4 32-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i64scatter_ epi64 Experimental Stores 4 64-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i64scatter_ pd Experimental Stores 4 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠i64scatter_ ps Experimental Stores 4 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm256_
mask_ ⚠insertf32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠insertf64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠inserti32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed 32-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠inserti64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed 64-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm256_
mask_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm256_
mask_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm256_
mask_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm256_
mask_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
mask_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using writemaskk
(elements are copied fromsrc
if the corresponding bit is not set). - _mm256_
mask_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex_ epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutex_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠permutexvar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm256_
mask_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm256_
mask_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm256_
mask_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm256_
mask_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
mask_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
mask_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
mask_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
mask_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using writemaskk
(elements are copied fromsrc
when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
mask_ ⚠reduce_ add_ epi8 Experimental Reduce the packed 8-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm256_
mask_ ⚠reduce_ add_ epi16 Experimental Reduce the packed 16-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm256_
mask_ ⚠reduce_ and_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise AND using mask k. Returns the bitwise AND of all active elements in a. - _mm256_
mask_ ⚠reduce_ and_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise AND using mask k. Returns the bitwise AND of all active elements in a. - _mm256_
mask_ ⚠reduce_ max_ epi8 Experimental Reduce the packed 8-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm256_
mask_ ⚠reduce_ max_ epi16 Experimental Reduce the packed 16-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm256_
mask_ ⚠reduce_ max_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm256_
mask_ ⚠reduce_ max_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm256_
mask_ ⚠reduce_ min_ epi8 Experimental Reduce the packed 8-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm256_
mask_ ⚠reduce_ min_ epi16 Experimental Reduce the packed 16-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm256_
mask_ ⚠reduce_ min_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm256_
mask_ ⚠reduce_ min_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm256_
mask_ ⚠reduce_ mul_ epi8 Experimental Reduce the packed 8-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm256_
mask_ ⚠reduce_ mul_ epi16 Experimental Reduce the packed 16-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm256_
mask_ ⚠reduce_ or_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm256_
mask_ ⚠reduce_ or_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm256_
mask_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
mask_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
mask_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
mask_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
mask_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
mask_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
mask_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
mask_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠set1_ epi16 Experimental Broadcast 16-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using writemask k (elements are copied from src“ when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ epi8 Experimental Shuffle 8-bit integers in a within 128-bit lanes using the control in the corresponding 8-bit element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠store_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠store_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠store_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠store_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
mask_ ⚠storeu_ epi8 Experimental Store packed 8-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠storeu_ epi16 Experimental Store packed 16-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠storeu_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠storeu_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠storeu_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠storeu_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm256_
mask_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 32-bit granularity (32-bit elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 64-bit granularity (64-bit elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠test_ epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm256_
mask_ ⚠test_ epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm256_
mask_ ⚠test_ epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm256_
mask_ ⚠test_ epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm256_
mask_ ⚠testn_ epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm256_
mask_ ⚠testn_ epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm256_
mask_ ⚠testn_ epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm256_
mask_ ⚠testn_ epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm256_
mask_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mask_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
mask_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm256_
maskz_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠abs_ epi32 Experimental Compute the absolute value of packed signed 32-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 32 bytes (8 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 64-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 32 bytes (4 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠and_ epi32 Experimental Compute the bitwise AND of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcast_ f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠broadcast_ f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcast_ f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠broadcast_ i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcast_ i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcastsd_ pd Experimental Broadcast the low double-precision (64-bit) floating-point element from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
maskz_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm256_
maskz_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
maskz_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm256_
maskz_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
maskz_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
maskz_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
maskz_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in the low 8 bytes of a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
maskz_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in single vector dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm256_
maskz_ ⚠cvtneps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm256_
maskz_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm256_
maskz_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
maskz_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm256_
maskz_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm256_
maskz_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm256_
maskz_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠extractf32x4_ ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠extractf64x2_ pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠extracti32x4_ epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM1, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠extracti64x2_ epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
maskz_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm256_
maskz_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm256_
maskz_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm256_
maskz_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
maskz_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
maskz_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
maskz_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm256_
maskz_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm256_
maskz_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
maskz_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm256_
maskz_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm256_
maskz_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
maskz_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm256_
maskz_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm256_
maskz_ ⚠insertf32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠insertf64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠inserti32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed 32-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠inserti64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed 64-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
maskz_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
maskz_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
maskz_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
maskz_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm256_
maskz_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm256_
maskz_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm256_
maskz_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
maskz_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using zeromaskk
(elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex_ epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutex_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠permutexvar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm256_
maskz_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm256_
maskz_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm256_
maskz_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm256_
maskz_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
maskz_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
maskz_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
maskz_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
maskz_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using zeromaskk
(elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
maskz_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
maskz_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
maskz_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
maskz_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
maskz_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
maskz_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm256_
maskz_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
maskz_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠set1_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ epi8 Experimental Shuffle packed 8-bit integers in a according to shuffle control mask in the corresponding 8-bit element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 32-bit granularity (32-bit elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 64-bit granularity (64-bit elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm256_
maskz_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
maskz_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm256_
max_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst. - _mm256_
max_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst. - _mm256_
max_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm256_
min_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst. - _mm256_
min_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst. - _mm256_
min_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm256_
mmask_ ⚠i32gather_ epi32 Experimental Loads 8 32-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i32gather_ epi64 Experimental Loads 4 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i32gather_ pd Experimental Loads 4 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i32gather_ ps Experimental Loads 8 single-precision (32-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i64gather_ epi32 Experimental Loads 4 32-bit integer elements from memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i64gather_ epi64 Experimental Loads 4 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i64gather_ pd Experimental Loads 4 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
mmask_ ⚠i64gather_ ps Experimental Loads 4 single-precision (32-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm256_
movepi8_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 8-bit integer in a. - _mm256_
movepi16_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 16-bit integer in a. - _mm256_
movepi32_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 32-bit integer in a. - _mm256_
movepi64_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 64-bit integer in a. - _mm256_
movm_ ⚠epi8 Experimental Set each packed 8-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm256_
movm_ ⚠epi16 Experimental Set each packed 16-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm256_
movm_ ⚠epi32 Experimental Set each packed 32-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm256_
movm_ ⚠epi64 Experimental Set each packed 64-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm256_
mul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm256_
mul_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm256_
mullo_ ⚠epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
. - _mm256_
multishift_ ⚠epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst. - _mm256_
or_ ⚠epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst. - _mm256_
or_ ⚠epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the resut in dst. - _mm256_
permutex2var_ ⚠epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a and b using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex2var_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm256_
permutex_ ⚠epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst. - _mm256_
permutex_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst. - _mm256_
permutexvar_ ⚠epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a using the corresponding index in idx, and store the results in dst. - _mm256_
permutexvar_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx. - _mm256_
popcnt_ ⚠epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm256_
popcnt_ ⚠epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm256_
popcnt_ ⚠epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm256_
popcnt_ ⚠epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm256_
range_ ⚠pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
range_ ⚠ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm256_
rcp14_ ⚠pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm256_
rcp14_ ⚠ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm256_
rcp_ ⚠ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
reduce_ ⚠add_ epi8 Experimental Reduce the packed 8-bit integers in a by addition. Returns the sum of all elements in a. - _mm256_
reduce_ ⚠add_ epi16 Experimental Reduce the packed 16-bit integers in a by addition. Returns the sum of all elements in a. - _mm256_
reduce_ ⚠add_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by addition. Returns the sum of all elements in a. - _mm256_
reduce_ ⚠and_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm256_
reduce_ ⚠and_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm256_
reduce_ ⚠max_ epi8 Experimental Reduce the packed 8-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm256_
reduce_ ⚠max_ epi16 Experimental Reduce the packed 16-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm256_
reduce_ ⚠max_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm256_
reduce_ ⚠max_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm256_
reduce_ ⚠max_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by maximum. Returns the maximum of all elements in a. - _mm256_
reduce_ ⚠min_ epi8 Experimental Reduce the packed 8-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm256_
reduce_ ⚠min_ epi16 Experimental Reduce the packed 16-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm256_
reduce_ ⚠min_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm256_
reduce_ ⚠min_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm256_
reduce_ ⚠min_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by minimum. Returns the minimum of all elements in a. - _mm256_
reduce_ ⚠mul_ epi8 Experimental Reduce the packed 8-bit integers in a by multiplication. Returns the product of all elements in a. - _mm256_
reduce_ ⚠mul_ epi16 Experimental Reduce the packed 16-bit integers in a by multiplication. Returns the product of all elements in a. - _mm256_
reduce_ ⚠mul_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by multiplication. Returns the product of all elements in a. - _mm256_
reduce_ ⚠or_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm256_
reduce_ ⚠or_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm256_
reduce_ ⚠pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
reduce_ ⚠ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. - _mm256_
reduce_ ⚠ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm256_
rol_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm256_
rol_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm256_
rolv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm256_
rolv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm256_
ror_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm256_
ror_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm256_
rorv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm256_
rorv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm256_
roundscale_ ⚠pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
roundscale_ ⚠ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst. - _mm256_
roundscale_ ⚠ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm256_
rsqrt14_ ⚠pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm256_
rsqrt14_ ⚠ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm256_
rsqrt_ ⚠ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm256_
scalef_ ⚠pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst. - _mm256_
scalef_ ⚠ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst. - _mm256_
scalef_ ⚠ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst. - _mm256_
set1_ ⚠ph Experimental Broadcast the half-precision (16-bit) floating-point value a to all elements of dst. - _mm256_
set_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values. - _mm256_
setr_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values in reverse order. - _mm256_
setzero_ ⚠ph Experimental Return vector of type __m256h with all elements set to zero. - _mm256_
shldi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst). - _mm256_
shldi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst. - _mm256_
shldi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst). - _mm256_
shldv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst. - _mm256_
shldv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst. - _mm256_
shldv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst. - _mm256_
shrdi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst. - _mm256_
shrdi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst. - _mm256_
shrdi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst. - _mm256_
shrdv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst. - _mm256_
shrdv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst. - _mm256_
shrdv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst. - _mm256_
shuffle_ ⚠f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst. - _mm256_
shuffle_ ⚠f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst. - _mm256_
shuffle_ ⚠i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst. - _mm256_
shuffle_ ⚠i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst. - _mm256_
sllv_ ⚠epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm256_
sqrt_ ⚠ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. - _mm256_
sra_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst. - _mm256_
srai_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst. - _mm256_
srav_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm256_
srav_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm256_
srlv_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm256_
store_ ⚠epi32 Experimental Store 256-bits (composed of 8 packed 32-bit integers) from a into memory. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
store_ ⚠epi64 Experimental Store 256-bits (composed of 4 packed 64-bit integers) from a into memory. mem_addr must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_
store_ ⚠ph Experimental Store 256-bits (composed of 16 packed half-precision (16-bit) floating-point elements) from a into memory. The address must be aligned to 32 bytes or a general-protection exception may be generated. - _mm256_
storeu_ ⚠epi8 Experimental Store 256-bits (composed of 32 packed 8-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm256_
storeu_ ⚠epi16 Experimental Store 256-bits (composed of 16 packed 16-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm256_
storeu_ ⚠epi32 Experimental Store 256-bits (composed of 8 packed 32-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm256_
storeu_ ⚠epi64 Experimental Store 256-bits (composed of 4 packed 64-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm256_
storeu_ ⚠ph Experimental Store 256-bits (composed of 16 packed half-precision (16-bit) floating-point elements) from a into memory. The address does not need to be aligned to any particular boundary. - _mm256_
sub_ ⚠ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst. - _mm256_
ternarylogic_ ⚠epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm256_
ternarylogic_ ⚠epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm256_
test_ ⚠epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm256_
test_ ⚠epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm256_
test_ ⚠epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm256_
test_ ⚠epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm256_
testn_ ⚠epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm256_
testn_ ⚠epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm256_
testn_ ⚠epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm256_
testn_ ⚠epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm256_
undefined_ ⚠ph Experimental Return vector of type__m256h
with undefined elements. In practice, this returns the all-zero vector. - _mm256_
xor_ ⚠epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst. - _mm256_
xor_ ⚠epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst. - _mm256_
zextph128_ ⚠ph256 Experimental Cast vector of type__m256h
to type__m128h
. The upper 8 elements of the result are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm512_
abs_ ⚠epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst. - _mm512_
abs_ ⚠epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst. - _mm512_
abs_ ⚠epi32 Experimental Computes the absolute values of packed 32-bit integers ina
. - _mm512_
abs_ ⚠epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst. - _mm512_
abs_ ⚠pd Experimental Finds the absolute value of each packed double-precision (64-bit) floating-point element in v2, storing the results in dst. - _mm512_
abs_ ⚠ph Experimental Finds the absolute value of each packed half-precision (16-bit) floating-point element in v2, storing the result in dst. - _mm512_
abs_ ⚠ps Experimental Finds the absolute value of each packed single-precision (32-bit) floating-point element in v2, storing the results in dst. - _mm512_
add_ ⚠epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst. - _mm512_
add_ ⚠epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst. - _mm512_
add_ ⚠epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst. - _mm512_
add_ ⚠epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst. - _mm512_
add_ ⚠pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
add_ ⚠ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
add_ ⚠ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
add_ ⚠round_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst.\ - _mm512_
add_ ⚠round_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
add_ ⚠round_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst.\ - _mm512_
adds_ ⚠epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst. - _mm512_
adds_ ⚠epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst. - _mm512_
adds_ ⚠epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst. - _mm512_
adds_ ⚠epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst. - _mm512_
aesdec_ ⚠epi128 Experimental Performs one round of an AES decryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm512_
aesdeclast_ ⚠epi128 Experimental Performs the last round of an AES decryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm512_
aesenc_ ⚠epi128 Experimental Performs one round of an AES encryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm512_
aesenclast_ ⚠epi128 Experimental Performs the last round of an AES encryption flow on each 128-bit word (state) ina
using the corresponding 128-bit word (key) inround_key
. - _mm512_
alignr_ ⚠epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst. - _mm512_
alignr_ ⚠epi32 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 64 bytes (16 elements) in dst. - _mm512_
alignr_ ⚠epi64 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 64 bytes (8 elements) in dst. - _mm512_
and_ ⚠epi32 Experimental Compute the bitwise AND of packed 32-bit integers in a and b, and store the results in dst. - _mm512_
and_ ⚠epi64 Experimental Compute the bitwise AND of 512 bits (composed of packed 64-bit integers) in a and b, and store the results in dst. - _mm512_
and_ ⚠pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst. - _mm512_
and_ ⚠ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst. - _mm512_
and_ ⚠si512 Experimental Compute the bitwise AND of 512 bits (representing integer data) in a and b, and store the result in dst. - _mm512_
andnot_ ⚠epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst. - _mm512_
andnot_ ⚠epi64 Experimental Compute the bitwise NOT of 512 bits (composed of packed 64-bit integers) in a and then AND with b, and store the results in dst. - _mm512_
andnot_ ⚠pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst. - _mm512_
andnot_ ⚠ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst. - _mm512_
andnot_ ⚠si512 Experimental Compute the bitwise NOT of 512 bits (representing integer data) in a and then AND with b, and store the result in dst. - _mm512_
avg_ ⚠epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst. - _mm512_
avg_ ⚠epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst. - _mm512_
bitshuffle_ ⚠epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm512_
broadcast_ ⚠f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst. - _mm512_
broadcast_ ⚠f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst. - _mm512_
broadcast_ ⚠f32x8 Experimental Broadcasts the 8 packed single-precision (32-bit) floating-point elements from a to all elements of dst. - _mm512_
broadcast_ ⚠f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst. - _mm512_
broadcast_ ⚠f64x4 Experimental Broadcast the 4 packed double-precision (64-bit) floating-point elements from a to all elements of dst. - _mm512_
broadcast_ ⚠i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst. - _mm512_
broadcast_ ⚠i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst. - _mm512_
broadcast_ ⚠i32x8 Experimental Broadcasts the 8 packed 32-bit integers from a to all elements of dst. - _mm512_
broadcast_ ⚠i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst. - _mm512_
broadcast_ ⚠i64x4 Experimental Broadcast the 4 packed 64-bit integers from a to all elements of dst. - _mm512_
broadcastb_ ⚠epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst. - _mm512_
broadcastd_ ⚠epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst. - _mm512_
broadcastmb_ ⚠epi64 Experimental Broadcast the low 8-bits from input mask k to all 64-bit elements of dst. - _mm512_
broadcastmw_ ⚠epi32 Experimental Broadcast the low 16-bits from input mask k to all 32-bit elements of dst. - _mm512_
broadcastq_ ⚠epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst. - _mm512_
broadcastsd_ ⚠pd Experimental Broadcast the low double-precision (64-bit) floating-point element from a to all elements of dst. - _mm512_
broadcastss_ ⚠ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst. - _mm512_
broadcastw_ ⚠epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst. - _mm512_
bslli_ ⚠epi128 Experimental Shift 128-bit lanes in a left by imm8 bytes while shifting in zeros, and store the results in dst. - _mm512_
bsrli_ ⚠epi128 Experimental Shift 128-bit lanes in a right by imm8 bytes while shifting in zeros, and store the results in dst. - _mm512_
castpd128_ ⚠pd512 Experimental Cast vector of type __m128d to type __m512d; the upper 384 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd256_ ⚠pd512 Experimental Cast vector of type __m256d to type __m512d; the upper 256 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd512_ ⚠pd128 Experimental Cast vector of type __m512d to type __m128d. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd512_ ⚠pd256 Experimental Cast vector of type __m512d to type __m256d. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd_ ⚠ph Experimental Cast vector of type__m512d
to type__m512h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd_ ⚠ps Experimental Cast vector of type __m512d to type __m512. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castpd_ ⚠si512 Experimental Cast vector of type __m512d to type __m512i. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castph128_ ⚠ph512 Experimental Cast vector of type__m128h
to type__m512h
. The upper 24 elements of the result are undefined. In practice, the upper elements are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm512_
castph256_ ⚠ph512 Experimental Cast vector of type__m256h
to type__m512h
. The upper 16 elements of the result are undefined. In practice, the upper elements are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm512_
castph512_ ⚠ph128 Experimental Cast vector of type__m512h
to type__m128h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castph512_ ⚠ph256 Experimental Cast vector of type__m512h
to type__m256h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castph_ ⚠pd Experimental Cast vector of type__m512h
to type__m512d
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castph_ ⚠ps Experimental Cast vector of type__m512h
to type__m512
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castph_ ⚠si512 Experimental Cast vector of type__m512h
to type__m512i
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps128_ ⚠ps512 Experimental Cast vector of type __m128 to type __m512; the upper 384 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps256_ ⚠ps512 Experimental Cast vector of type __m256 to type __m512; the upper 256 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps512_ ⚠ps128 Experimental Cast vector of type __m512 to type __m128. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps512_ ⚠ps256 Experimental Cast vector of type __m512 to type __m256. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps_ ⚠pd Experimental Cast vector of type __m512 to type __m512d. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps_ ⚠ph Experimental Cast vector of type__m512
to type__m512h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castps_ ⚠si512 Experimental Cast vector of type __m512 to type __m512i. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi128_ ⚠si512 Experimental Cast vector of type __m128i to type __m512i; the upper 384 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi256_ ⚠si512 Experimental Cast vector of type __m256i to type __m512i; the upper 256 bits of the result are undefined. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi512_ ⚠pd Experimental Cast vector of type __m512i to type __m512d. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi512_ ⚠ph Experimental Cast vector of type__m512i
to type__m512h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi512_ ⚠ps Experimental Cast vector of type __m512i to type __m512. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi512_ ⚠si128 Experimental Cast vector of type __m512i to type __m128i. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
castsi512_ ⚠si256 Experimental Cast vector of type __m512i to type __m256i. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
clmulepi64_ ⚠epi128 Experimental Performs a carry-less multiplication of two 64-bit polynomials over the finite field GF(2) - in each of the 4 128-bit lanes. - _mm512_
cmp_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified byIMM8
, and store the results in mask vector k. - _mm512_
cmp_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠round_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cmp_ ⚠round_ ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm512_
cmp_ ⚠round_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cmpeq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for equality, and store the results in mask vector k. - _mm512_
cmpeq_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for equality, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpge_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmpgt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm512_
cmple_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmple_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for less-than-or-equal, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for less-than, and store the results in mask vector k. - _mm512_
cmplt_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for less-than, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpneq_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-equal, and store the results in mask vector k. - _mm512_
cmpnle_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-less-than-or-equal, and store the results in mask vector k. - _mm512_
cmpnle_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-less-than-or-equal, and store the results in mask vector k. - _mm512_
cmpnlt_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-less-than, and store the results in mask vector k. - _mm512_
cmpnlt_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-less-than, and store the results in mask vector k. - _mm512_
cmpord_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b to see if neither is NaN, and store the results in mask vector k. - _mm512_
cmpord_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b to see if neither is NaN, and store the results in mask vector k. - _mm512_
cmpunord_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b to see if either is NaN, and store the results in mask vector k. - _mm512_
cmpunord_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b to see if either is NaN, and store the results in mask vector k. - _mm512_
cmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
cmul_ ⚠round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
conflict_ ⚠epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
conflict_ ⚠epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
conj_ ⚠pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
cvt_ ⚠roundepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst.\ - _mm512_
cvt_ ⚠roundepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepu32_ ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst.\ - _mm512_
cvt_ ⚠roundepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst.\ - _mm512_
cvt_ ⚠roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst.\ - _mm512_
cvt_ ⚠roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst.\ - _mm512_
cvt_ ⚠roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvt_ ⚠roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvt_ ⚠roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvt_ ⚠roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst.\ - _mm512_
cvt_ ⚠roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
cvt_ ⚠roundps_ pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvt_ ⚠roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtepi8_ ⚠epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst. - _mm512_
cvtepi8_ ⚠epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtepi8_ ⚠epi64 Experimental Sign extend packed 8-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepi16_ ⚠epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi16_ ⚠epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtepi16_ ⚠epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepi16_ ⚠ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi32_ ⚠epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi32_ ⚠epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi32_ ⚠epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepi32_ ⚠pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi32_ ⚠ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi32_ ⚠ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi32lo_ ⚠pd Experimental Performs element-by-element conversion of the lower half of packed 32-bit integer elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst. - _mm512_
cvtepi64_ ⚠epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi64_ ⚠epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi64_ ⚠epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst. - _mm512_
cvtepi64_ ⚠pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi64_ ⚠ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepi64_ ⚠ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu8_ ⚠epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst. - _mm512_
cvtepu8_ ⚠epi32 Experimental Zero extend packed unsigned 8-bit integers in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtepu8_ ⚠epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 8 byte sof a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepu16_ ⚠epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtepu16_ ⚠epi64 Experimental Zero extend packed unsigned 16-bit integers in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepu16_ ⚠ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu32_ ⚠epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtepu32_ ⚠pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu32_ ⚠ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu32_ ⚠ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu32lo_ ⚠pd Experimental Performs element-by-element conversion of the lower half of packed 32-bit unsigned integer elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst. - _mm512_
cvtepu64_ ⚠pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu64_ ⚠ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtepu64_ ⚠ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtne2ps_ ⚠pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two 512-bit vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in a 512-bit wide vector. Intel’s documentation - _mm512_
cvtneps_ ⚠pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst. Intel’s documentation - _mm512_
cvtpbh_ ⚠ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtpd_ ⚠epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm512_
cvtpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm512_
cvtpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm512_
cvtpd_ ⚠ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtpd_ ⚠ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtpd_ ⚠pslo Experimental Performs an element-by-element conversion of packed double-precision (64-bit) floating-point elements in v2 to single-precision (32-bit) floating-point elements and stores them in dst. The elements are stored in the lower half of the results vector, while the remaining upper half locations are set to 0. - _mm512_
cvtph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst. - _mm512_
cvtph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst. - _mm512_
cvtph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst. - _mm512_
cvtph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst. - _mm512_
cvtph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst. - _mm512_
cvtph_ ⚠pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtps_ ⚠epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm512_
cvtps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm512_
cvtps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm512_
cvtps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm512_
cvtps_ ⚠pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm512_
cvtps_ ⚠ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtpslo_ ⚠pd Experimental Performs element-by-element conversion of the lower half of packed single-precision (32-bit) floating-point elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst. - _mm512_
cvtsd_ ⚠f64 Experimental Copy the lower double-precision (64-bit) floating-point element of a to dst. - _mm512_
cvtsepi16_ ⚠epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsepi32_ ⚠epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsepi32_ ⚠epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsepi64_ ⚠epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsepi64_ ⚠epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsepi64_ ⚠epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst. - _mm512_
cvtsh_ ⚠h Experimental Copy the lower half-precision (16-bit) floating-point element froma
todst
. - _mm512_
cvtsi512_ ⚠si32 Experimental Copy the lower 32-bit integer in a to dst. - _mm512_
cvtss_ ⚠f32 Experimental Copy the lower single-precision (32-bit) floating-point element of a to dst. - _mm512_
cvtt_ ⚠roundpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtt_ ⚠roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
cvtt_ ⚠roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtt_ ⚠roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
cvtt_ ⚠roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst. - _mm512_
cvtt_ ⚠roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtt_ ⚠roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
cvtt_ ⚠roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
cvtt_ ⚠roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
cvttpd_ ⚠epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm512_
cvttpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm512_
cvttpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm512_
cvttpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm512_
cvttph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst. - _mm512_
cvttph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm512_
cvttph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst. - _mm512_
cvttph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst. - _mm512_
cvttph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst. - _mm512_
cvttph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst. - _mm512_
cvttps_ ⚠epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm512_
cvttps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm512_
cvttps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm512_
cvttps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm512_
cvtusepi16_ ⚠epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtusepi32_ ⚠epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtusepi32_ ⚠epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtusepi64_ ⚠epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtusepi64_ ⚠epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtusepi64_ ⚠epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst. - _mm512_
cvtx_ ⚠roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtx_ ⚠roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
cvtxph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm512_
cvtxps_ ⚠ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm512_
dbsad_ ⚠epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst. Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm512_
div_ ⚠pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst. - _mm512_
div_ ⚠ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst. - _mm512_
div_ ⚠ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst. - _mm512_
div_ ⚠round_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, =and store the results in dst.\ - _mm512_
div_ ⚠round_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
div_ ⚠round_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst.\ - _mm512_
dpbf16_ ⚠ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst.Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst. Intel’s documentation - _mm512_
dpbusd_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm512_
dpbusds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm512_
dpwssd_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm512_
dpwssds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm512_
extractf32x4_ ⚠ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the result in dst. - _mm512_
extractf32x8_ ⚠ps Experimental Extracts 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst. - _mm512_
extractf64x2_ ⚠pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst. - _mm512_
extractf64x4_ ⚠pd Experimental Extract 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from a, selected with imm8, and store the result in dst. - _mm512_
extracti32x4_ ⚠epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM2, and store the result in dst. - _mm512_
extracti32x8_ ⚠epi32 Experimental Extracts 256 bits (composed of 8 packed 32-bit integers) from a, selected with IMM8, and stores the result in dst. - _mm512_
extracti64x2_ ⚠epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst. - _mm512_
extracti64x4_ ⚠epi64 Experimental Extract 256 bits (composed of 4 packed 64-bit integers) from a, selected with IMM1, and store the result in dst. - _mm512_
fcmadd_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
fcmadd_ ⚠round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
fcmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
fcmul_ ⚠round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm512_
fixupimm_ ⚠pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm512_
fixupimm_ ⚠ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm512_
fixupimm_ ⚠round_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting.\ - _mm512_
fixupimm_ ⚠round_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting.\ - _mm512_
fmadd_ ⚠pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
fmadd_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm512_
fmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm512_
fmadd_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm512_
fmadd_ ⚠round_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
fmadd_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst.\ - _mm512_
fmadd_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm512_
fmadd_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst.\ - _mm512_
fmaddsub_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmaddsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmaddsub_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmaddsub_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst.\ - _mm512_
fmaddsub_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmaddsub_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst.\ - _mm512_
fmsub_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. - _mm512_
fmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. - _mm512_
fmsub_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. - _mm512_
fmsub_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst.\ - _mm512_
fmsub_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. - _mm512_
fmsub_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst.\ - _mm512_
fmsubadd_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst. - _mm512_
fmsubadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmsubadd_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst. - _mm512_
fmsubadd_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst.\ - _mm512_
fmsubadd_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst. - _mm512_
fmsubadd_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst.\ - _mm512_
fmul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
fmul_ ⚠round_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
fnmadd_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst. - _mm512_
fnmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst. - _mm512_
fnmadd_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst. - _mm512_
fnmadd_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst.\ - _mm512_
fnmadd_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst. - _mm512_
fnmadd_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst.\ - _mm512_
fnmsub_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm512_
fnmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm512_
fnmsub_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm512_
fnmsub_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst.\ - _mm512_
fnmsub_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm512_
fnmsub_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst.\ - _mm512_
fpclass_ ⚠pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm512_
fpclass_ ⚠ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm512_
fpclass_ ⚠ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm512_
getexp_ ⚠pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
getexp_ ⚠ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm512_
getexp_ ⚠ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
getexp_ ⚠round_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
getexp_ ⚠round_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculatesfloor(log2(x))
for each element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
getexp_ ⚠round_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
getmant_ ⚠pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
getmant_ ⚠ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm512_
getmant_ ⚠ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign. The mantissa is normalized to the interval specified by interv, which can take the following values: _MM_MANT_NORM_1_2 // interval [1, 2) _MM_MANT_NORM_p5_2 // interval [0.5, 2) _MM_MANT_NORM_p5_1 // interval [0.5, 1) _MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5) The sign is determined by sc which can take the following values: _MM_MANT_SIGN_src // sign = sign(src) _MM_MANT_SIGN_zero // sign = 0 _MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
getmant_ ⚠round_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
getmant_ ⚠round_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
getmant_ ⚠round_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
gf2p8affine_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
gf2p8affineinv_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
gf2p8mul_ ⚠epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm512_
i32gather_ ⚠epi32 Experimental Gather 32-bit integers from memory using 32-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i32gather_ ⚠epi64 Experimental Gather 64-bit integers from memory using 32-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i32gather_ ⚠pd Experimental Gather double-precision (64-bit) floating-point elements from memory using 32-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i32gather_ ⚠ps Experimental Gather single-precision (32-bit) floating-point elements from memory using 32-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i32logather_ ⚠epi64 Experimental Loads 8 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale and stores them in dst. - _mm512_
i32logather_ ⚠pd Experimental Loads 8 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale and stores them in dst. - _mm512_
i32loscatter_ ⚠epi64 Experimental Stores 8 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale. - _mm512_
i32loscatter_ ⚠pd Experimental Stores 8 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale. - _mm512_
i32scatter_ ⚠epi32 Experimental Scatter 32-bit integers from a into memory using 32-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i32scatter_ ⚠epi64 Experimental Scatter 64-bit integers from a into memory using 32-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i32scatter_ ⚠pd Experimental Scatter double-precision (64-bit) floating-point elements from a into memory using 32-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i32scatter_ ⚠ps Experimental Scatter single-precision (32-bit) floating-point elements from a into memory using 32-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i64gather_ ⚠epi32 Experimental Gather 32-bit integers from memory using 64-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i64gather_ ⚠epi64 Experimental Gather 64-bit integers from memory using 64-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i64gather_ ⚠pd Experimental Gather double-precision (64-bit) floating-point elements from memory using 64-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i64gather_ ⚠ps Experimental Gather single-precision (32-bit) floating-point elements from memory using 64-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst. scale should be 1, 2, 4 or 8. - _mm512_
i64scatter_ ⚠epi32 Experimental Scatter 32-bit integers from a into memory using 64-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i64scatter_ ⚠epi64 Experimental Scatter 64-bit integers from a into memory using 64-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i64scatter_ ⚠pd Experimental Scatter double-precision (64-bit) floating-point elements from a into memory using 64-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). scale should be 1, 2, 4 or 8. - _mm512_
i64scatter_ ⚠ps Experimental Scatter single-precision (32-bit) floating-point elements from a into memory using 64-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
insertf32x4 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into dst at the location specified by imm8. - _mm512_
insertf32x8 ⚠Experimental Copy a to dst, then insert 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from b into dst at the location specified by IMM8. - _mm512_
insertf64x2 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into dst at the location specified by IMM8. - _mm512_
insertf64x4 ⚠Experimental Copy a to dst, then insert 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from b into dst at the location specified by imm8. - _mm512_
inserti32x4 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 4 packed 32-bit integers) from b into dst at the location specified by imm8. - _mm512_
inserti32x8 ⚠Experimental Copy a to dst, then insert 256 bits (composed of 8 packed 32-bit integers) from b into dst at the location specified by IMM8. - _mm512_
inserti64x2 ⚠Experimental Copy a to dst, then insert 128 bits (composed of 2 packed 64-bit integers) from b into dst at the location specified by IMM8. - _mm512_
inserti64x4 ⚠Experimental Copy a to dst, then insert 256 bits (composed of 4 packed 64-bit integers) from b into dst at the location specified by imm8. - _mm512_
int2mask ⚠Experimental Converts integer mask into bitmask, storing the result in dst. - _mm512_
kand ⚠Experimental Compute the bitwise AND of 16-bit masks a and b, and store the result in k. - _mm512_
kandn ⚠Experimental Compute the bitwise NOT of 16-bit masks a and then AND with b, and store the result in k. - _mm512_
kmov ⚠Experimental Copy 16-bit mask a to k. - _mm512_
knot ⚠Experimental Compute the bitwise NOT of 16-bit mask a, and store the result in k. - _mm512_
kor ⚠Experimental Compute the bitwise OR of 16-bit masks a and b, and store the result in k. - _mm512_
kortestc ⚠Experimental Performs bitwise OR between k1 and k2, storing the result in dst. CF flag is set if dst consists of all 1’s. - _mm512_
kortestz ⚠Experimental Performs bitwise OR between k1 and k2, storing the result in dst. ZF flag is set if dst is 0. - _mm512_
kunpackb ⚠Experimental Unpack and interleave 8 bits from masks a and b, and store the 16-bit result in k. - _mm512_
kunpackd ⚠Experimental Unpack and interleave 32 bits from masks a and b, and store the 64-bit result in k. - _mm512_
kunpackw ⚠Experimental Unpack and interleave 16 bits from masks a and b, and store the 32-bit result in k. - _mm512_
kxnor ⚠Experimental Compute the bitwise XNOR of 16-bit masks a and b, and store the result in k. - _mm512_
kxor ⚠Experimental Compute the bitwise XOR of 16-bit masks a and b, and store the result in k. - _mm512_
load_ ⚠epi32 Experimental Load 512-bits (composed of 16 packed 32-bit integers) from memory into dst. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
load_ ⚠epi64 Experimental Load 512-bits (composed of 8 packed 64-bit integers) from memory into dst. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
load_ ⚠pd Experimental Load 512-bits (composed of 8 packed double-precision (64-bit) floating-point elements) from memory into dst. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
load_ ⚠ph Experimental Load 512-bits (composed of 32 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address must be aligned to 64 bytes or a general-protection exception may be generated. - _mm512_
load_ ⚠ps Experimental Load 512-bits (composed of 16 packed single-precision (32-bit) floating-point elements) from memory into dst. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
load_ ⚠si512 Experimental Load 512-bits of integer data from memory into dst. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
loadu_ ⚠epi8 Experimental Load 512-bits (composed of 64 packed 8-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠epi16 Experimental Load 512-bits (composed of 32 packed 16-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠epi32 Experimental Load 512-bits (composed of 16 packed 32-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠epi64 Experimental Load 512-bits (composed of 8 packed 64-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠pd Experimental Loads 512-bits (composed of 8 packed double-precision (64-bit) floating-point elements) from memory into result.mem_addr
does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠ph Experimental Load 512-bits (composed of 32 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address does not need to be aligned to any particular boundary. - _mm512_
loadu_ ⚠ps Experimental Loads 512-bits (composed of 16 packed single-precision (32-bit) floating-point elements) from memory into result.mem_addr
does not need to be aligned on any particular boundary. - _mm512_
loadu_ ⚠si512 Experimental Load 512-bits of integer data from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm512_
lzcnt_ ⚠epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst. - _mm512_
lzcnt_ ⚠epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst. - _mm512_
madd52hi_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm512_
madd52lo_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm512_
madd_ ⚠epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst. - _mm512_
maddubs_ ⚠epi16 Experimental Vertically multiply each unsigned 8-bit integer from a with the corresponding signed 8-bit integer from b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst. - _mm512_
mask2_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask2_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm512_
mask2_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm512_
mask2_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm512_
mask2_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set) - _mm512_
mask2_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm512_
mask2int ⚠Experimental Converts bit mask k1 into an integer value, storing the results in dst. - _mm512_
mask3_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask3_ ⚠fcmadd_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c using writemask k (the element is copied from c when the corresponding mask bit is not set), and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask3_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask3_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmadd_ round_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask3_ ⚠fmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmaddsub_ pd Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmaddsub_ round_ pd Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmaddsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmaddsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsubadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fmsubadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fmsubadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fnmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask3_ ⚠fnmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm512_
mask3_ ⚠fnmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠abs_ epi32 Experimental Computes the absolute value of packed 32-bit integers ina
, and store the unsigned results indst
using writemaskk
(elements are copied fromsrc
when the corresponding mask bit is not set). - _mm512_
mask_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠abs_ pd Experimental Finds the absolute value of each packed double-precision (64-bit) floating-point element in v2, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠abs_ ps Experimental Finds the absolute value of each packed single-precision (32-bit) floating-point element in v2, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠add_ round_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠add_ round_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠add_ round_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 64 bytes (16 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 64 bytes (8 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠and_ epi32 Experimental Performs element-by-element bitwise AND between packed 32-bit integer elements of a and b, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠bitshuffle_ epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm512_
mask_ ⚠blend_ epi8 Experimental Blend packed 8-bit integers from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ epi16 Experimental Blend packed 16-bit integers from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ epi32 Experimental Blend packed 32-bit integers from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ epi64 Experimental Blend packed 64-bit integers from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ pd Experimental Blend packed double-precision (64-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ ph Experimental Blend packed half-precision (16-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠blend_ ps Experimental Blend packed single-precision (32-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm512_
mask_ ⚠broadcast_ f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcast_ f32x8 Experimental Broadcasts the 8 packed single-precision (32-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ f64x4 Experimental Broadcast the 4 packed double-precision (64-bit) floating-point elements from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcast_ i32x8 Experimental Broadcasts the 8 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠broadcast_ i64x4 Experimental Broadcast the 4 packed 64-bit integers from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastsd_ pd Experimental Broadcast the low double-precision (64-bit) floating-point element from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ round_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cmp_ round_ ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmp_ round_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cmpeq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpeq_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpge_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpgt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmple_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmplt_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpneq_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpnle_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpnle_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpnlt_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b for not-less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpnlt_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b for not-less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpord_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b to see if neither is NaN, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpord_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b to see if neither is NaN, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpunord_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b to see if either is NaN, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmpunord_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b to see if either is NaN, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠cmul_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm512_
mask_ ⚠compressstoreu_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠compressstoreu_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠compressstoreu_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠compressstoreu_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠compressstoreu_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠compressstoreu_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
mask_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
mask_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠cvt_ roundepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepu32_ ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvt_ roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvt_ roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠cvt_ roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
mask_ ⚠cvt_ roundps_ pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi16_ storeu_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi32_ storeu_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepi32_ storeu_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepi32lo_ pd Experimental Performs element-by-element conversion of the lower half of packed 32-bit integer elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtepi64_ storeu_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepi64_ storeu_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepi64_ storeu_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu32_ ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu32lo_ pd Experimental Performs element-by-element conversion of the lower half of 32-bit unsigned integer elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in single vector dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm512_
mask_ ⚠cvtneps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm512_
mask_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtpd_ pslo Experimental Performs an element-by-element conversion of packed double-precision (64-bit) floating-point elements in v2 to single-precision (32-bit) floating-point elements and stores them in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The elements are stored in the lower half of the results vector, while the remaining upper half locations are set to 0. - _mm512_
mask_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtps_ pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtpslo_ pd Experimental Performs element-by-element conversion of the lower half of packed single-precision (32-bit) floating-point elements in v2 to packed double-precision (64-bit) floating-point elements, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi16_ storeu_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi32_ storeu_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtsepi32_ storeu_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtsepi64_ storeu_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtsepi64_ storeu_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtsepi64_ storeu_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtt_ roundpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtt_ roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠cvtt_ roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
mask_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi16_ storeu_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi32_ storeu_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtusepi32_ storeu_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtusepi64_ storeu_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtusepi64_ storeu_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtusepi64_ storeu_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed 32-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm512_
mask_ ⚠cvtx_ roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtx_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm512_
mask_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm512_
mask_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠div_ round_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠div_ round_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠div_ round_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm512_
mask_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠extractf32x4_ ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠extractf32x8_ ps Experimental Extracts 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠extractf64x2_ pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠extractf64x4_ pd Experimental Extract 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from a, selected with imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠extracti32x4_ epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM2, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠extracti32x8_ epi32 Experimental Extracts 256 bits (composed of 8 packed 32-bit integers) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠extracti64x2_ epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠extracti64x4_ epi64 Experimental Extract 256 bits (composed of 4 packed 64-bit integers) from a, selected with IMM1, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠fcmadd_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠fcmul_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
mask_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm512_
mask_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm512_
mask_ ⚠fixupimm_ round_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting.\ - _mm512_
mask_ ⚠fixupimm_ round_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting.\ - _mm512_
mask_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmadd_ round_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask_ ⚠fmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmaddsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmaddsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmaddsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsubadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmsubadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fmsubadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask_ ⚠fmul_ round_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fnmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fnmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠fnmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠fpclass_ pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm512_
mask_ ⚠fpclass_ ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm512_
mask_ ⚠fpclass_ ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm512_
mask_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
mask_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm512_
mask_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
mask_ ⚠getexp_ round_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠getexp_ round_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
mask_ ⚠getexp_ round_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
mask_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm512_
mask_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
mask_ ⚠getmant_ round_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠getmant_ round_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
mask_ ⚠getmant_ round_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
mask_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
mask_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm512_
mask_ ⚠i32gather_ epi32 Experimental Gather 32-bit integers from memory using 32-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32gather_ epi64 Experimental Gather 64-bit integers from memory using 32-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32gather_ pd Experimental Gather double-precision (64-bit) floating-point elements from memory using 32-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32gather_ ps Experimental Gather single-precision (32-bit) floating-point elements from memory using 32-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32logather_ epi64 Experimental Loads 8 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale and stores them in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠i32logather_ pd Experimental Loads 8 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale and stores them in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠i32loscatter_ epi64 Experimental Stores 8 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm512_
mask_ ⚠i32loscatter_ pd Experimental Stores 8 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in the lower half of vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm512_
mask_ ⚠i32scatter_ epi32 Experimental Scatter 32-bit integers from a into memory using 32-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32scatter_ epi64 Experimental Scatter 64-bit integers from a into memory using 32-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32scatter_ pd Experimental Scatter double-precision (64-bit) floating-point elements from a into memory using 32-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i32scatter_ ps Experimental Scatter single-precision (32-bit) floating-point elements from a into memory using 32-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 32-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64gather_ epi32 Experimental Gather 32-bit integers from memory using 64-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64gather_ epi64 Experimental Gather 64-bit integers from memory using 64-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64gather_ pd Experimental Gather double-precision (64-bit) floating-point elements from memory using 64-bit indices. 64-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64gather_ ps Experimental Gather single-precision (32-bit) floating-point elements from memory using 64-bit indices. 32-bit elements are loaded from addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale). Gathered elements are merged into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64scatter_ epi32 Experimental Scatter 32-bit integers from a into memory using 64-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64scatter_ epi64 Experimental Scatter 64-bit integers from a into memory using 64-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64scatter_ pd Experimental Scatter double-precision (64-bit) floating-point elements from a into memory using 64-bit indices. 64-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠i64scatter_ ps Experimental Scatter single-precision (32-bit) floating-point elements from a into memory using 64-bit indices. 32-bit elements are stored at addresses starting at base_addr and offset by each 64-bit element in vindex (each index is scaled by the factor in scale) subject to mask k (elements are not stored when the corresponding mask bit is not set). scale should be 1, 2, 4 or 8. - _mm512_
mask_ ⚠insertf32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠insertf32x8 Experimental Copy a to tmp, then insert 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠insertf64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠insertf64x4 Experimental Copy a to tmp, then insert 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠inserti32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed 32-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠inserti32x8 Experimental Copy a to tmp, then insert 256 bits (composed of 8 packed 32-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠inserti64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed 64-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠inserti64x4 Experimental Copy a to tmp, then insert 256 bits (composed of 4 packed 64-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm512_
mask_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm512_
mask_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
mask_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠max_ round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠max_ round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
mask_ ⚠max_ round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
mask_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠min_ round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠min_ round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
mask_ ⚠min_ round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mul_ round_ pch Experimental Multiply the packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mask_ ⚠mul_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠mul_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠mul_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using writemaskk
(elements are copied fromsrc
if the corresponding bit is not set). - _mm512_
mask_ ⚠mullox_ epi64 Experimental Multiplies elements in packed 64-bit integer vectors a and b together, storing the lower 64 bits of the result in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutevar_ epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Note that this intrinsic shuffles across 128-bit lanes, unlike past intrinsics that use the permutevar name. This intrinsic is identical to _mm512_mask_permutexvar_epi32, and it is recommended that you use that intrinsic name. - _mm512_
mask_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex_ epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutex_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠permutexvar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm512_
mask_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm512_
mask_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm512_
mask_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm512_
mask_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
mask_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
mask_ ⚠range_ round_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
mask_ ⚠range_ round_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
mask_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
mask_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
mask_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using writemaskk
(elements are copied fromsrc
when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
mask_ ⚠reduce_ add_ epi32 Experimental Reduce the packed 32-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm512_
mask_ ⚠reduce_ add_ epi64 Experimental Reduce the packed 64-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm512_
mask_ ⚠reduce_ add_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by addition using mask k. Returns the sum of all active elements in a. - _mm512_
mask_ ⚠reduce_ add_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by addition using mask k. Returns the sum of all active elements in a. - _mm512_
mask_ ⚠reduce_ and_ epi32 Experimental Reduce the packed 32-bit integers in a by bitwise AND using mask k. Returns the bitwise AND of all active elements in a. - _mm512_
mask_ ⚠reduce_ and_ epi64 Experimental Reduce the packed 64-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ epi32 Experimental Reduce the packed signed 32-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ epi64 Experimental Reduce the packed signed 64-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ epu32 Experimental Reduce the packed unsigned 32-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ epu64 Experimental Reduce the packed unsigned 64-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ max_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ epi32 Experimental Reduce the packed signed 32-bit integers in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ epi64 Experimental Reduce the packed signed 64-bit integers in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ epu32 Experimental Reduce the packed unsigned 32-bit integers in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ epu64 Experimental Reduce the packed signed 64-bit integers in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ min_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by maximum using mask k. Returns the minimum of all active elements in a. - _mm512_
mask_ ⚠reduce_ mul_ epi32 Experimental Reduce the packed 32-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm512_
mask_ ⚠reduce_ mul_ epi64 Experimental Reduce the packed 64-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm512_
mask_ ⚠reduce_ mul_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by multiplication using mask k. Returns the product of all active elements in a. - _mm512_
mask_ ⚠reduce_ mul_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by multiplication using mask k. Returns the product of all active elements in a. - _mm512_
mask_ ⚠reduce_ or_ epi32 Experimental Reduce the packed 32-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm512_
mask_ ⚠reduce_ or_ epi64 Experimental Reduce the packed 64-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm512_
mask_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
mask_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
mask_ ⚠reduce_ round_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
mask_ ⚠reduce_ round_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠reduce_ round_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
mask_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
mask_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
mask_ ⚠roundscale_ round_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
mask_ ⚠roundscale_ round_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
mask_ ⚠roundscale_ round_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
mask_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
mask_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
mask_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
mask_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠scalef_ round_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠scalef_ round_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠scalef_ round_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠set1_ epi16 Experimental Broadcast 16-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using writemask k (elements are copied from src“ when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ epi8 Experimental Shuffle 8-bit integers in a within 128-bit lanes using the control in the corresponding 8-bit element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sqrt_ round_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠sqrt_ round_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠sqrt_ round_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠store_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠store_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠store_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠store_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
mask_ ⚠storeu_ epi8 Experimental Store packed 8-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠storeu_ epi16 Experimental Store packed 16-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠storeu_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠storeu_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠storeu_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠storeu_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm512_
mask_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠sub_ round_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠sub_ round_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mask_ ⚠sub_ round_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).\ - _mm512_
mask_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 32-bit granularity (32-bit elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 64-bit granularity (64-bit elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠test_ epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm512_
mask_ ⚠test_ epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm512_
mask_ ⚠test_ epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm512_
mask_ ⚠test_ epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm512_
mask_ ⚠testn_ epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm512_
mask_ ⚠testn_ epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm512_
mask_ ⚠testn_ epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm512_
mask_ ⚠testn_ epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm512_
mask_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm512_
mask_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
mask_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm512_
maskz_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠abs_ epi32 Experimental Computes the absolute value of packed 32-bit integers ina
, and store the unsigned results indst
using zeromaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠add_ round_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠add_ round_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠add_ round_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 32-bit elements, and stores the low 64 bytes (16 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 128-byte immediate result, shift the result right by imm8 64-bit elements, and stores the low 64 bytes (8 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠and_ epi32 Experimental Compute the bitwise AND of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcast_ f32x2 Experimental Broadcasts the lower 2 packed single-precision (32-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ f32x4 Experimental Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcast_ f32x8 Experimental Broadcasts the 8 packed single-precision (32-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ f64x2 Experimental Broadcasts the 2 packed double-precision (64-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ f64x4 Experimental Broadcast the 4 packed double-precision (64-bit) floating-point elements from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ i32x4 Experimental Broadcast the 4 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcast_ i32x8 Experimental Broadcasts the 8 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ i64x2 Experimental Broadcasts the 2 packed 64-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠broadcast_ i64x4 Experimental Broadcast the 4 packed 64-bit integers from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastsd_ pd Experimental Broadcast the low double-precision (64-bit) floating-point element from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠cmul_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm512_
maskz_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
maskz_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm512_
maskz_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠cvt_ roundepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepu32_ ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvt_ roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvt_ roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠cvt_ roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Rounding is done according to the ROUNDING parameter, which can be one of: - _mm512_
maskz_ ⚠cvt_ roundps_ pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 8 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu32_ ps Experimental Convert packed unsigned 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in single vector dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm512_
maskz_ ⚠cvtneps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm512_
maskz_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtps_ pd Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm512_
maskz_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundpd_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtt_ roundps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠cvtt_ roundps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC to the sae parameter. - _mm512_
maskz_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding - _mm512_
maskz_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtx_ roundph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtx_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm512_
maskz_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠div_ round_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠div_ round_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠div_ round_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm512_
maskz_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠extractf32x4_ ps Experimental Extract 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from a, selected with imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠extractf32x8_ ps Experimental Extracts 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠extractf64x2_ pd Experimental Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠extractf64x4_ pd Experimental Extract 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from a, selected with imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠extracti32x4_ epi32 Experimental Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM2, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠extracti32x8_ epi32 Experimental Extracts 256 bits (composed of 8 packed 32-bit integers) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠extracti64x2_ epi64 Experimental Extracts 128 bits (composed of 2 packed 64-bit integers) from a, selected with IMM8, and stores the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠extracti64x4_ epi64 Experimental Extract 256 bits (composed of 4 packed 64-bit integers) from a, selected with IMM1, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fcmadd_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c using zeromask k (the element is zeroed out when the corresponding mask bit is not set), and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fcmul_ round_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm512_
maskz_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm512_
maskz_ ⚠fixupimm_ round_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting.\ - _mm512_
maskz_ ⚠fixupimm_ round_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting.\ - _mm512_
maskz_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmadd_ round_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in a using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmaddsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmaddsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmaddsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsubadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmsubadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fmsubadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
maskz_ ⚠fmul_ round_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmadd_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fnmadd_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmadd_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmsub_ round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠fnmsub_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠fnmsub_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
maskz_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm512_
maskz_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm512_
maskz_ ⚠getexp_ round_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠getexp_ round_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
maskz_ ⚠getexp_ round_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
maskz_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm512_
maskz_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm512_
maskz_ ⚠getmant_ round_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠getmant_ round_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
maskz_ ⚠getmant_ round_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
maskz_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm512_
maskz_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm512_
maskz_ ⚠insertf32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠insertf32x8 Experimental Copy a to tmp, then insert 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠insertf64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠insertf64x4 Experimental Copy a to tmp, then insert 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠inserti32x4 Experimental Copy a to tmp, then insert 128 bits (composed of 4 packed 32-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠inserti32x8 Experimental Copy a to tmp, then insert 256 bits (composed of 8 packed 32-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠inserti64x2 Experimental Copy a to tmp, then insert 128 bits (composed of 2 packed 64-bit integers) from b into tmp at the location specified by IMM8, and copy tmp to dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠inserti64x4 Experimental Copy a to tmp, then insert 256 bits (composed of 4 packed 64-bit integers) from b into tmp at the location specified by imm8. Store tmp to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
maskz_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
maskz_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
maskz_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
maskz_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm512_
maskz_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
maskz_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠max_ round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠max_ round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
maskz_ ⚠max_ round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
maskz_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠min_ round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠min_ round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
maskz_ ⚠min_ round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
maskz_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mul_ round_ pch Experimental Multiply the packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
maskz_ ⚠mul_ round_ pd Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠mul_ round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠mul_ round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using zeromaskk
(elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex_ epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutex_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠permutexvar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm512_
maskz_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm512_
maskz_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm512_
maskz_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm512_
maskz_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
maskz_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
maskz_ ⚠range_ round_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
maskz_ ⚠range_ round_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
maskz_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
maskz_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
maskz_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using zeromaskk
(elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
maskz_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
maskz_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
maskz_ ⚠reduce_ round_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
maskz_ ⚠reduce_ round_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠reduce_ round_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
maskz_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
maskz_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
maskz_ ⚠roundscale_ round_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
maskz_ ⚠roundscale_ round_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
maskz_ ⚠roundscale_ round_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
maskz_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
maskz_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm512_
maskz_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
maskz_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠scalef_ round_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠scalef_ round_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠scalef_ round_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠set1_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ epi8 Experimental Shuffle packed 8-bit integers in a according to shuffle control mask in the corresponding 8-bit element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sqrt_ round_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠sqrt_ round_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠sqrt_ round_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠sub_ round_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠sub_ round_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm512_
maskz_ ⚠sub_ round_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).\ - _mm512_
maskz_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 32-bit granularity (32-bit elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 64-bit granularity (64-bit elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm512_
maskz_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
maskz_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm512_
max_ ⚠epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
max_ ⚠ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst. - _mm512_
max_ ⚠round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
max_ ⚠round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm512_
max_ ⚠round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
min_ ⚠epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst. Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
min_ ⚠ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst. - _mm512_
min_ ⚠round_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
min_ ⚠round_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm512_
min_ ⚠round_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
movedup_ ⚠pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst. - _mm512_
movehdup_ ⚠ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst. - _mm512_
moveldup_ ⚠ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst. - _mm512_
movepi8_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 8-bit integer in a. - _mm512_
movepi16_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 16-bit integer in a. - _mm512_
movepi32_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 32-bit integer in a. - _mm512_
movepi64_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 64-bit integer in a. - _mm512_
movm_ ⚠epi8 Experimental Set each packed 8-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm512_
movm_ ⚠epi16 Experimental Set each packed 16-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm512_
movm_ ⚠epi32 Experimental Set each packed 32-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm512_
movm_ ⚠epi64 Experimental Set each packed 64-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm512_
mul_ ⚠epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst. - _mm512_
mul_ ⚠epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst. - _mm512_
mul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mul_ ⚠pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
mul_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
mul_ ⚠ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. - _mm512_
mul_ ⚠round_ pch Experimental Multiply the packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm512_
mul_ ⚠round_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst.\ - _mm512_
mul_ ⚠round_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
mul_ ⚠round_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst.\ - _mm512_
mulhi_ ⚠epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst. - _mm512_
mulhi_ ⚠epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst. - _mm512_
mulhrs_ ⚠epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst. - _mm512_
mullo_ ⚠epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst. - _mm512_
mullo_ ⚠epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst. - _mm512_
mullo_ ⚠epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
. - _mm512_
mullox_ ⚠epi64 Experimental Multiplies elements in packed 64-bit integer vectors a and b together, storing the lower 64 bits of the result in dst. - _mm512_
multishift_ ⚠epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst. - _mm512_
or_ ⚠epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst. - _mm512_
or_ ⚠epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the resut in dst. - _mm512_
or_ ⚠pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst. - _mm512_
or_ ⚠ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst. - _mm512_
or_ ⚠si512 Experimental Compute the bitwise OR of 512 bits (representing integer data) in a and b, and store the result in dst. - _mm512_
packs_ ⚠epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst. - _mm512_
packs_ ⚠epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst. - _mm512_
packus_ ⚠epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst. - _mm512_
packus_ ⚠epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst. - _mm512_
permute_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst. - _mm512_
permute_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst. - _mm512_
permutevar_ ⚠epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. Note that this intrinsic shuffles across 128-bit lanes, unlike past intrinsics that use the permutevar name. This intrinsic is identical to _mm512_permutexvar_epi32, and it is recommended that you use that intrinsic name. - _mm512_
permutevar_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst. - _mm512_
permutevar_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst. - _mm512_
permutex2var_ ⚠epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a and b using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex2var_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm512_
permutex_ ⚠epi64 Experimental Shuffle 64-bit integers in a within 256-bit lanes using the control in imm8, and store the results in dst. - _mm512_
permutex_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 256-bit lanes using the control in imm8, and store the results in dst. - _mm512_
permutexvar_ ⚠epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠epi32 Experimental Shuffle 32-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠epi64 Experimental Shuffle 64-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a across lanes using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a using the corresponding index in idx, and store the results in dst. - _mm512_
permutexvar_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a across lanes using the corresponding index in idx. - _mm512_
popcnt_ ⚠epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm512_
popcnt_ ⚠epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm512_
popcnt_ ⚠epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm512_
popcnt_ ⚠epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm512_
range_ ⚠pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
range_ ⚠ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm512_
range_ ⚠round_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
range_ ⚠round_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm512_
rcp14_ ⚠pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm512_
rcp14_ ⚠ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm512_
rcp_ ⚠ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
reduce_ ⚠add_ epi32 Experimental Reduce the packed 32-bit integers in a by addition. Returns the sum of all elements in a. - _mm512_
reduce_ ⚠add_ epi64 Experimental Reduce the packed 64-bit integers in a by addition. Returns the sum of all elements in a. - _mm512_
reduce_ ⚠add_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by addition. Returns the sum of all elements in a. - _mm512_
reduce_ ⚠add_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by addition. Returns the sum of all elements in a. - _mm512_
reduce_ ⚠add_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by addition. Returns the sum of all elements in a. - _mm512_
reduce_ ⚠and_ epi32 Experimental Reduce the packed 32-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm512_
reduce_ ⚠and_ epi64 Experimental Reduce the packed 64-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm512_
reduce_ ⚠max_ epi32 Experimental Reduce the packed signed 32-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ epi64 Experimental Reduce the packed signed 64-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ epu32 Experimental Reduce the packed unsigned 32-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ epu64 Experimental Reduce the packed unsigned 64-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠max_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by maximum. Returns the maximum of all elements in a. - _mm512_
reduce_ ⚠min_ epi32 Experimental Reduce the packed signed 32-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ epi64 Experimental Reduce the packed signed 64-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ epu32 Experimental Reduce the packed unsigned 32-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ epu64 Experimental Reduce the packed unsigned 64-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠min_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by minimum. Returns the minimum of all elements in a. - _mm512_
reduce_ ⚠mul_ epi32 Experimental Reduce the packed 32-bit integers in a by multiplication. Returns the product of all elements in a. - _mm512_
reduce_ ⚠mul_ epi64 Experimental Reduce the packed 64-bit integers in a by multiplication. Returns the product of all elements in a. - _mm512_
reduce_ ⚠mul_ pd Experimental Reduce the packed double-precision (64-bit) floating-point elements in a by multiplication. Returns the product of all elements in a. - _mm512_
reduce_ ⚠mul_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by multiplication. Returns the product of all elements in a. - _mm512_
reduce_ ⚠mul_ ps Experimental Reduce the packed single-precision (32-bit) floating-point elements in a by multiplication. Returns the product of all elements in a. - _mm512_
reduce_ ⚠or_ epi32 Experimental Reduce the packed 32-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm512_
reduce_ ⚠or_ epi64 Experimental Reduce the packed 64-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm512_
reduce_ ⚠pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
reduce_ ⚠ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. - _mm512_
reduce_ ⚠ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
reduce_ ⚠round_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
reduce_ ⚠round_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. - _mm512_
reduce_ ⚠round_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm512_
rol_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm512_
rol_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm512_
rolv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm512_
rolv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm512_
ror_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm512_
ror_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm512_
rorv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm512_
rorv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm512_
roundscale_ ⚠pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
roundscale_ ⚠ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst. - _mm512_
roundscale_ ⚠ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm512_
roundscale_ ⚠round_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
roundscale_ ⚠round_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm512_
roundscale_ ⚠round_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm512_
rsqrt14_ ⚠pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm512_
rsqrt14_ ⚠ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm512_
rsqrt_ ⚠ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm512_
sad_ ⚠epu8 Experimental Compute the absolute differences of packed unsigned 8-bit integers in a and b, then horizontally sum each consecutive 8 differences to produce eight unsigned 16-bit integers, and pack these unsigned 16-bit integers in the low 16 bits of 64-bit elements in dst. - _mm512_
scalef_ ⚠pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst. - _mm512_
scalef_ ⚠ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst. - _mm512_
scalef_ ⚠ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst. - _mm512_
scalef_ ⚠round_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst.\ - _mm512_
scalef_ ⚠round_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst. - _mm512_
scalef_ ⚠round_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst.\ - _mm512_
set1_ ⚠epi8 Experimental Broadcast 8-bit integer a to all elements of dst. - _mm512_
set1_ ⚠epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst. - _mm512_
set1_ ⚠epi32 Experimental Broadcast 32-bit integera
to all elements ofdst
. - _mm512_
set1_ ⚠epi64 Experimental Broadcast 64-bit integera
to all elements ofdst
. - _mm512_
set1_ ⚠pd Experimental Broadcast 64-bit floata
to all elements ofdst
. - _mm512_
set1_ ⚠ph Experimental Broadcast the half-precision (16-bit) floating-point value a to all elements of dst. - _mm512_
set1_ ⚠ps Experimental Broadcast 32-bit floata
to all elements ofdst
. - _mm512_
set4_ ⚠epi32 Experimental Set packed 32-bit integers in dst with the repeated 4 element sequence. - _mm512_
set4_ ⚠epi64 Experimental Set packed 64-bit integers in dst with the repeated 4 element sequence. - _mm512_
set4_ ⚠pd Experimental Set packed double-precision (64-bit) floating-point elements in dst with the repeated 4 element sequence. - _mm512_
set4_ ⚠ps Experimental Set packed single-precision (32-bit) floating-point elements in dst with the repeated 4 element sequence. - _mm512_
set_ ⚠epi8 Experimental Set packed 8-bit integers in dst with the supplied values. - _mm512_
set_ ⚠epi16 Experimental Set packed 16-bit integers in dst with the supplied values. - _mm512_
set_ ⚠epi32 Experimental Sets packed 32-bit integers indst
with the supplied values. - _mm512_
set_ ⚠epi64 Experimental Set packed 64-bit integers in dst with the supplied values. - _mm512_
set_ ⚠pd Experimental Set packed double-precision (64-bit) floating-point elements in dst with the supplied values. - _mm512_
set_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values. - _mm512_
set_ ⚠ps Experimental Sets packed 32-bit integers indst
with the supplied values. - _mm512_
setr4_ ⚠epi32 Experimental Set packed 32-bit integers in dst with the repeated 4 element sequence in reverse order. - _mm512_
setr4_ ⚠epi64 Experimental Set packed 64-bit integers in dst with the repeated 4 element sequence in reverse order. - _mm512_
setr4_ ⚠pd Experimental Set packed double-precision (64-bit) floating-point elements in dst with the repeated 4 element sequence in reverse order. - _mm512_
setr4_ ⚠ps Experimental Set packed single-precision (32-bit) floating-point elements in dst with the repeated 4 element sequence in reverse order. - _mm512_
setr_ ⚠epi32 Experimental Sets packed 32-bit integers indst
with the supplied values in reverse order. - _mm512_
setr_ ⚠epi64 Experimental Set packed 64-bit integers in dst with the supplied values in reverse order. - _mm512_
setr_ ⚠pd Experimental Set packed double-precision (64-bit) floating-point elements in dst with the supplied values in reverse order. - _mm512_
setr_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values in reverse order. - _mm512_
setr_ ⚠ps Experimental Sets packed 32-bit integers indst
with the supplied values in reverse order. - _mm512_
setzero ⚠Experimental Return vector of type__m512
with all elements set to zero. - _mm512_
setzero_ ⚠epi32 Experimental Return vector of type__m512i
with all elements set to zero. - _mm512_
setzero_ ⚠pd Experimental Returns vector of type__m512d
with all elements set to zero. - _mm512_
setzero_ ⚠ph Experimental Return vector of type __m512h with all elements set to zero. - _mm512_
setzero_ ⚠ps Experimental Returns vector of type__m512
with all elements set to zero. - _mm512_
setzero_ ⚠si512 Experimental Returns vector of type__m512i
with all elements set to zero. - _mm512_
shldi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst). - _mm512_
shldi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst. - _mm512_
shldi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst). - _mm512_
shldv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst. - _mm512_
shldv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst. - _mm512_
shldv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst. - _mm512_
shrdi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst. - _mm512_
shrdi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst. - _mm512_
shrdi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst. - _mm512_
shrdv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst. - _mm512_
shrdv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst. - _mm512_
shrdv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst. - _mm512_
shuffle_ ⚠epi8 Experimental Shuffle packed 8-bit integers in a according to shuffle control mask in the corresponding 8-bit element of b, and store the results in dst. - _mm512_
shuffle_ ⚠epi32 Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst. - _mm512_
shuffle_ ⚠f32x4 Experimental Shuffle 128-bits (composed of 4 single-precision (32-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst. - _mm512_
shuffle_ ⚠f64x2 Experimental Shuffle 128-bits (composed of 2 double-precision (64-bit) floating-point elements) selected by imm8 from a and b, and store the results in dst. - _mm512_
shuffle_ ⚠i32x4 Experimental Shuffle 128-bits (composed of 4 32-bit integers) selected by imm8 from a and b, and store the results in dst. - _mm512_
shuffle_ ⚠i64x2 Experimental Shuffle 128-bits (composed of 2 64-bit integers) selected by imm8 from a and b, and store the results in dst. - _mm512_
shuffle_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst. - _mm512_
shuffle_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst. - _mm512_
shufflehi_ ⚠epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst. - _mm512_
shufflelo_ ⚠epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst. - _mm512_
sll_ ⚠epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst. - _mm512_
sll_ ⚠epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst. - _mm512_
sll_ ⚠epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst. - _mm512_
slli_ ⚠epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst. - _mm512_
slli_ ⚠epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst. - _mm512_
slli_ ⚠epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst. - _mm512_
sllv_ ⚠epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
sllv_ ⚠epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
sllv_ ⚠epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
sqrt_ ⚠pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. - _mm512_
sqrt_ ⚠ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. - _mm512_
sqrt_ ⚠ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. - _mm512_
sqrt_ ⚠round_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst.\ - _mm512_
sqrt_ ⚠round_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
sqrt_ ⚠round_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst.\ - _mm512_
sra_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst. - _mm512_
sra_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst. - _mm512_
sra_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst. - _mm512_
srai_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst. - _mm512_
srai_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst. - _mm512_
srai_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst. - _mm512_
srav_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm512_
srav_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm512_
srav_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm512_
srl_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst. - _mm512_
srl_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst. - _mm512_
srl_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst. - _mm512_
srli_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst. - _mm512_
srli_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst. - _mm512_
srli_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst. - _mm512_
srlv_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
srlv_ ⚠epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
srlv_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm512_
store_ ⚠epi32 Experimental Store 512-bits (composed of 16 packed 32-bit integers) from a into memory. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
store_ ⚠epi64 Experimental Store 512-bits (composed of 8 packed 64-bit integers) from a into memory. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
store_ ⚠pd Experimental Store 512-bits (composed of 8 packed double-precision (64-bit) floating-point elements) from a into memory. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
store_ ⚠ph Experimental Store 512-bits (composed of 32 packed half-precision (16-bit) floating-point elements) from a into memory. The address must be aligned to 64 bytes or a general-protection exception may be generated. - _mm512_
store_ ⚠ps Experimental Store 512-bits of integer data from a into memory. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
store_ ⚠si512 Experimental Store 512-bits of integer data from a into memory. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
storeu_ ⚠epi8 Experimental Store 512-bits (composed of 64 packed 8-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠epi16 Experimental Store 512-bits (composed of 32 packed 16-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠epi32 Experimental Store 512-bits (composed of 16 packed 32-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠epi64 Experimental Store 512-bits (composed of 8 packed 64-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠pd Experimental Stores 512-bits (composed of 8 packed double-precision (64-bit) floating-point elements) froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠ph Experimental Store 512-bits (composed of 32 packed half-precision (16-bit) floating-point elements) from a into memory. The address does not need to be aligned to any particular boundary. - _mm512_
storeu_ ⚠ps Experimental Stores 512-bits (composed of 16 packed single-precision (32-bit) floating-point elements) froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm512_
storeu_ ⚠si512 Experimental Store 512-bits of integer data from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm512_
stream_ ⚠load_ si512 Experimental Load 512-bits of integer data from memory into dst using a non-temporal memory hint. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. To minimize caching, the data is flagged as non-temporal (unlikely to be used again soon) - _mm512_
stream_ ⚠pd Experimental Store 512-bits (composed of 8 packed double-precision (64-bit) floating-point elements) from a into memory using a non-temporal memory hint. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
stream_ ⚠ps Experimental Store 512-bits (composed of 16 packed single-precision (32-bit) floating-point elements) from a into memory using a non-temporal memory hint. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
stream_ ⚠si512 Experimental Store 512-bits of integer data from a into memory using a non-temporal memory hint. mem_addr must be aligned on a 64-byte boundary or a general-protection exception may be generated. - _mm512_
sub_ ⚠epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst. - _mm512_
sub_ ⚠epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst. - _mm512_
sub_ ⚠epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst. - _mm512_
sub_ ⚠epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst. - _mm512_
sub_ ⚠pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst. - _mm512_
sub_ ⚠ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst. - _mm512_
sub_ ⚠ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst. - _mm512_
sub_ ⚠round_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst.\ - _mm512_
sub_ ⚠round_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst. Rounding is done according to the rounding parameter, which can be one of: - _mm512_
sub_ ⚠round_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst.\ - _mm512_
subs_ ⚠epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst. - _mm512_
subs_ ⚠epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst. - _mm512_
subs_ ⚠epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst. - _mm512_
subs_ ⚠epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst. - _mm512_
ternarylogic_ ⚠epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm512_
ternarylogic_ ⚠epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm512_
test_ ⚠epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm512_
test_ ⚠epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm512_
test_ ⚠epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm512_
test_ ⚠epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm512_
testn_ ⚠epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm512_
testn_ ⚠epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm512_
testn_ ⚠epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm512_
testn_ ⚠epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm512_
undefined ⚠Experimental Return vector of type __m512 with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm512_
undefined_ ⚠epi32 Experimental Return vector of type __m512i with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm512_
undefined_ ⚠pd Experimental Returns vector of type__m512d
with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm512_
undefined_ ⚠ph Experimental Return vector of type__m512h
with undefined elements. In practice, this returns the all-zero vector. - _mm512_
undefined_ ⚠ps Experimental Returns vector of type__m512
with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm512_
unpackhi_ ⚠epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpackhi_ ⚠epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpackhi_ ⚠epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpackhi_ ⚠epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpackhi_ ⚠pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpackhi_ ⚠ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
unpacklo_ ⚠ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst. - _mm512_
xor_ ⚠epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst. - _mm512_
xor_ ⚠epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst. - _mm512_
xor_ ⚠pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst. - _mm512_
xor_ ⚠ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst. - _mm512_
xor_ ⚠si512 Experimental Compute the bitwise XOR of 512 bits (representing integer data) in a and b, and store the result in dst. - _mm512_
zextpd128_ ⚠pd512 Experimental Cast vector of type __m128d to type __m512d; the upper 384 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
zextpd256_ ⚠pd512 Experimental Cast vector of type __m256d to type __m512d; the upper 256 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
zextph128_ ⚠ph512 Experimental Cast vector of type__m128h
to type__m512h
. The upper 24 elements of the result are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm512_
zextph256_ ⚠ph512 Experimental Cast vector of type__m256h
to type__m512h
. The upper 16 elements of the result are zeroed. This intrinsic can generate thevzeroupper
instruction, but most of the time it does not generate any instructions. - _mm512_
zextps128_ ⚠ps512 Experimental Cast vector of type __m128 to type __m512; the upper 384 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
zextps256_ ⚠ps512 Experimental Cast vector of type __m256 to type __m512; the upper 256 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
zextsi128_ ⚠si512 Experimental Cast vector of type __m128i to type __m512i; the upper 384 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm512_
zextsi256_ ⚠si512 Experimental Cast vector of type __m256i to type __m512i; the upper 256 bits of the result are zeroed. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
abs_ ⚠epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst. - _mm_
abs_ ⚠ph Experimental Finds the absolute value of each packed half-precision (16-bit) floating-point element in v2, storing the results in dst. - _mm_
add_ ⚠ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm_
add_ ⚠round_ sd Experimental Add the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
add_ ⚠round_ sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
add_ ⚠round_ ss Experimental Add the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
add_ ⚠sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
alignr_ ⚠epi32 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 16 bytes (4 elements) in dst. - _mm_
alignr_ ⚠epi64 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 16 bytes (2 elements) in dst. - _mm_
bcstnebf16_ ⚠ps Experimental Convert scalar BF16 (16-bit) floating point element stored at memory locations starting at location a to single precision (32-bit) floating-point, broadcast it to packed single precision (32-bit) floating-point elements, and store the results in dst. - _mm_
bcstnesh_ ⚠ps Experimental Convert scalar half-precision (16-bit) floating-point element stored at memory locations starting at location a to a single-precision (32-bit) floating-point, broadcast it to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm_
bitshuffle_ ⚠epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm_
broadcast_ ⚠i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst. - _mm_
broadcastmb_ ⚠epi64 Experimental Broadcast the low 8-bits from input mask k to all 64-bit elements of dst. - _mm_
broadcastmw_ ⚠epi32 Experimental Broadcast the low 16-bits from input mask k to all 32-bit elements of dst. - _mm_
castpd_ ⚠ph Experimental Cast vector of type__m128d
to type__m128h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
castph_ ⚠pd Experimental Cast vector of type__m128h
to type__m128d
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
castph_ ⚠ps Experimental Cast vector of type__m128h
to type__m128
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
castph_ ⚠si128 Experimental Cast vector of type__m128h
to type__m128i
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
castps_ ⚠ph Experimental Cast vector of type__m128
to type__m128h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
castsi128_ ⚠ph Experimental Cast vector of type__m128i
to type__m128h
. This intrinsic is only used for compilation and does not generate any instructions, thus it has zero latency. - _mm_
cmp_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k. - _mm_
cmp_ ⚠round_ sd_ mask Experimental Compare the lower double-precision (64-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cmp_ ⚠round_ sh_ mask Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the result in mask vector k. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cmp_ ⚠round_ ss_ mask Experimental Compare the lower single-precision (32-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cmp_ ⚠sd_ mask Experimental Compare the lower double-precision (64-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k. - _mm_
cmp_ ⚠sh_ mask Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the result in mask vector k. - _mm_
cmp_ ⚠ss_ mask Experimental Compare the lower single-precision (32-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k. - _mm_
cmpeq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpeq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k. - _mm_
cmpge_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpge_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmpgt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k. - _mm_
cmple_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmple_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k. - _mm_
cmplt_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmplt_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmpneq_ ⚠epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k. - _mm_
cmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
cmul_ ⚠round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm_
cmul_ ⚠sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm_
comi_ ⚠round_ sd Experimental Compare the lower double-precision (64-bit) floating-point element in a and b based on the comparison operand specified by imm8, and return the boolean result (0 or 1).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
comi_ ⚠round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and return the boolean result (0 or 1). Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
comi_ ⚠round_ ss Experimental Compare the lower single-precision (32-bit) floating-point element in a and b based on the comparison operand specified by imm8, and return the boolean result (0 or 1).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
comi_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and return the boolean result (0 or 1). - _mm_
comieq_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for equality, and return the boolean result (0 or 1). - _mm_
comige_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for greater-than-or-equal, and return the boolean result (0 or 1). - _mm_
comigt_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for greater-than, and return the boolean result (0 or 1). - _mm_
comile_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for less-than-or-equal, and return the boolean result (0 or 1). - _mm_
comilt_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for less-than, and return the boolean result (0 or 1). - _mm_
comineq_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for not-equal, and return the boolean result (0 or 1). - _mm_
conflict_ ⚠epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm_
conflict_ ⚠epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit. Each element’s comparison forms a zero extended bit vector in dst. - _mm_
conj_ ⚠pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
cvt_ ⚠roundi32_ sh Experimental Convert the signed 32-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundi32_ ss Experimental Convert the signed 32-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
cvt_ ⚠roundi64_ sd Experimental Convert the signed 64-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundi64_ sh Experimental Convert the signed 64-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundi64_ ss Experimental Convert the signed 64-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ i32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ i64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundsd_ si32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ si64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ ss Experimental Convert the lower double-precision (64-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ u32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsd_ u64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsh_ i32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit integer, and store the result in dst. - _mm_
cvt_ ⚠roundsh_ i64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit integer, and store the result in dst. - _mm_
cvt_ ⚠roundsh_ sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvt_ ⚠roundsh_ ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundsh_ u32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit unsigned integer, and store the result in dst. - _mm_
cvt_ ⚠roundsh_ u64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit unsigned integer, and store the result in dst. - _mm_
cvt_ ⚠roundsi32_ ss Experimental Convert the signed 32-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
cvt_ ⚠roundsi64_ sd Experimental Convert the signed 64-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundsi64_ ss Experimental Convert the signed 64-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ i32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ i64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ sd Experimental Convert the lower single-precision (32-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvt_ ⚠roundss_ sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundss_ si32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ si64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ u32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 32-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundss_ u64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 64-bit integer, and store the result in dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundu32_ sh Experimental Convert the unsigned 32-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundu32_ ss Experimental Convert the unsigned 32-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundu64_ sd Experimental Convert the unsigned 64-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvt_ ⚠roundu64_ sh Experimental Convert the unsigned 64-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 1 packed elements from a to the upper elements of dst. - _mm_
cvt_ ⚠roundu64_ ss Experimental Convert the unsigned 64-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
cvtepi16_ ⚠epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm_
cvtepi16_ ⚠ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm_
cvtepi32_ ⚠epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm_
cvtepi32_ ⚠epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm_
cvtepi32_ ⚠ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 64 bits of dst are zeroed out. - _mm_
cvtepi64_ ⚠epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst. - _mm_
cvtepi64_ ⚠epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst. - _mm_
cvtepi64_ ⚠epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst. - _mm_
cvtepi64_ ⚠pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm_
cvtepi64_ ⚠ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 96 bits of dst are zeroed out. - _mm_
cvtepi64_ ⚠ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtepu16_ ⚠ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm_
cvtepu32_ ⚠pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm_
cvtepu32_ ⚠ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 64 bits of dst are zeroed out. - _mm_
cvtepu64_ ⚠pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm_
cvtepu64_ ⚠ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 96 bits of dst are zeroed out. - _mm_
cvtepu64_ ⚠ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvti32_ ⚠sd Experimental Convert the signed 32-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvti32_ ⚠sh Experimental Convert the signed 32-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvti32_ ⚠ss Experimental Convert the signed 32-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvti64_ ⚠sd Experimental Convert the signed 64-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvti64_ ⚠sh Experimental Convert the signed 64-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvti64_ ⚠ss Experimental Convert the signed 64-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvtne2ps_ ⚠pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two 128-bit vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in a 128-bit wide vector. Intel’s documentation - _mm_
cvtneebf16_ ⚠ps Experimental Convert packed BF16 (16-bit) floating-point even-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtneeph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point even-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtneobf16_ ⚠ps Experimental Convert packed BF16 (16-bit) floating-point odd-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtneoph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point odd-indexed elements stored at memory locations starting at location a to single precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtneps_ ⚠avx_ pbh Experimental Convert packed single precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst. - _mm_
cvtneps_ ⚠pbh Experimental Converts packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst. - _mm_
cvtness_ ⚠sbh Experimental Converts a single-precision (32-bit) floating-point element in a to a BF16 (16-bit) floating-point element, and store the result in dst. - _mm_
cvtpbh_ ⚠ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to single-precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm_
cvtpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm_
cvtpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm_
cvtpd_ ⚠ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. The upper 96 bits of dst are zeroed out. - _mm_
cvtph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst. - _mm_
cvtph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm_
cvtph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst. - _mm_
cvtph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst. - _mm_
cvtph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst. - _mm_
cvtph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst. - _mm_
cvtph_ ⚠pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst. - _mm_
cvtps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst. - _mm_
cvtps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst. - _mm_
cvtps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst. - _mm_
cvtsbh_ ⚠ss Experimental Converts a single BF16 (16-bit) floating-point element in a to a single-precision (32-bit) floating-point element, and store the result in dst. - _mm_
cvtsd_ ⚠i32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 32-bit integer, and store the result in dst. - _mm_
cvtsd_ ⚠i64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer, and store the result in dst. - _mm_
cvtsd_ ⚠sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvtsd_ ⚠u32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 32-bit integer, and store the result in dst. - _mm_
cvtsd_ ⚠u64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 64-bit integer, and store the result in dst. - _mm_
cvtsepi16_ ⚠epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsepi32_ ⚠epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsepi32_ ⚠epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsepi64_ ⚠epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsepi64_ ⚠epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsepi64_ ⚠epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst. - _mm_
cvtsh_ ⚠h Experimental Copy the lower half-precision (16-bit) floating-point element froma
todst
. - _mm_
cvtsh_ ⚠i32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit integer, and store the result in dst. - _mm_
cvtsh_ ⚠i64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit integer, and store the result in dst. - _mm_
cvtsh_ ⚠sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvtsh_ ⚠ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvtsh_ ⚠u32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit unsigned integer, and store the result in dst. - _mm_
cvtsh_ ⚠u64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit unsigned integer, and store the result in dst. - _mm_
cvtsi16_ ⚠si128 Experimental Copy 16-bit integer a to the lower elements of dst, and zero the upper elements of dst. - _mm_
cvtsi128_ ⚠si16 Experimental Copy the lower 16-bit integer in a to dst. - _mm_
cvtss_ ⚠i32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer, and store the result in dst. - _mm_
cvtss_ ⚠i64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer, and store the result in dst. - _mm_
cvtss_ ⚠sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvtss_ ⚠u32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 32-bit integer, and store the result in dst. - _mm_
cvtss_ ⚠u64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 64-bit integer, and store the result in dst. - _mm_
cvtt_ ⚠roundsd_ i32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsd_ i64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsd_ si32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsd_ si64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsd_ u32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsd_ u64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundsh_ i32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst. - _mm_
cvtt_ ⚠roundsh_ i64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst. - _mm_
cvtt_ ⚠roundsh_ u32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit unsigned integer with truncation, and store the result in dst. - _mm_
cvtt_ ⚠roundsh_ u64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit unsigned integer with truncation, and store the result in dst. - _mm_
cvtt_ ⚠roundss_ i32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundss_ i64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundss_ si32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundss_ si64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundss_ u32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 32-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvtt_ ⚠roundss_ u64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 64-bit integer with truncation, and store the result in dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
cvttpd_ ⚠epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm_
cvttpd_ ⚠epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm_
cvttpd_ ⚠epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm_
cvttph_ ⚠epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst. - _mm_
cvttph_ ⚠epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst. - _mm_
cvttph_ ⚠epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst. - _mm_
cvttph_ ⚠epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst. - _mm_
cvttph_ ⚠epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst. - _mm_
cvttph_ ⚠epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst. - _mm_
cvttps_ ⚠epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst. - _mm_
cvttps_ ⚠epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst. - _mm_
cvttps_ ⚠epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst. - _mm_
cvttsd_ ⚠i32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst. - _mm_
cvttsd_ ⚠i64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst. - _mm_
cvttsd_ ⚠u32 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 32-bit integer with truncation, and store the result in dst. - _mm_
cvttsd_ ⚠u64 Experimental Convert the lower double-precision (64-bit) floating-point element in a to an unsigned 64-bit integer with truncation, and store the result in dst. - _mm_
cvttsh_ ⚠i32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst. - _mm_
cvttsh_ ⚠i64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst. - _mm_
cvttsh_ ⚠u32 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 32-bit unsigned integer with truncation, and store the result in dst. - _mm_
cvttsh_ ⚠u64 Experimental Convert the lower half-precision (16-bit) floating-point element in a to a 64-bit unsigned integer with truncation, and store the result in dst. - _mm_
cvttss_ ⚠i32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 32-bit integer with truncation, and store the result in dst. - _mm_
cvttss_ ⚠i64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to a 64-bit integer with truncation, and store the result in dst. - _mm_
cvttss_ ⚠u32 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 32-bit integer with truncation, and store the result in dst. - _mm_
cvttss_ ⚠u64 Experimental Convert the lower single-precision (32-bit) floating-point element in a to an unsigned 64-bit integer with truncation, and store the result in dst. - _mm_
cvtu32_ ⚠sd Experimental Convert the unsigned 32-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvtu32_ ⚠sh Experimental Convert the unsigned 32-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
cvtu32_ ⚠ss Experimental Convert the unsigned 32-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvtu64_ ⚠sd Experimental Convert the unsigned 64-bit integer b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
cvtu64_ ⚠sh Experimental Convert the unsigned 64-bit integer b to a half-precision (16-bit) floating-point element, store the result in the lower element of dst, and copy the upper 1 packed elements from a to the upper elements of dst. - _mm_
cvtu64_ ⚠ss Experimental Convert the unsigned 64-bit integer b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
cvtusepi16_ ⚠epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtusepi32_ ⚠epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtusepi32_ ⚠epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtusepi64_ ⚠epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtusepi64_ ⚠epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtusepi64_ ⚠epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst. - _mm_
cvtxph_ ⚠ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst. - _mm_
cvtxps_ ⚠ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst. - _mm_
dbsad_ ⚠epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst. Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm_
div_ ⚠ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst. - _mm_
div_ ⚠round_ sd Experimental Divide the lower double-precision (64-bit) floating-point element in a by the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
div_ ⚠round_ sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
div_ ⚠round_ ss Experimental Divide the lower single-precision (32-bit) floating-point element in a by the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
div_ ⚠sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
dpbf16_ ⚠ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst. Intel’s documentation - _mm_
dpbssd_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpbssds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
dpbsud_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpbsuds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
dpbusd_ ⚠avx_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpbusd_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpbusds_ ⚠avx_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm_
dpbusds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm_
dpbuud_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpbuuds_ ⚠epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
dpwssd_ ⚠avx_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpwssd_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpwssds_ ⚠avx_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm_
dpwssds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst. - _mm_
dpwsud_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpwsuds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
dpwusd_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding signed 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpwusds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding signed 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
dpwuud_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst. - _mm_
dpwuuds_ ⚠epi32 Experimental Multiply groups of 2 adjacent pairs of unsigned 16-bit integers in a with corresponding unsigned 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src with signed saturation, and store the packed 32-bit results in dst. - _mm_
fcmadd_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
fcmadd_ ⚠round_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
fcmadd_ ⚠sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
fcmul_ ⚠pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
fcmul_ ⚠round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm_
fcmul_ ⚠sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
fixupimm_ ⚠pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm_
fixupimm_ ⚠ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst. imm8 is used to set the required flags reporting. - _mm_
fixupimm_ ⚠round_ sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
fixupimm_ ⚠round_ ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
fixupimm_ ⚠sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting. - _mm_
fixupimm_ ⚠ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting. - _mm_
fmadd_ ⚠pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst. - _mm_
fmadd_ ⚠round_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fmadd_ ⚠round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
fmadd_ ⚠round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fmadd_ ⚠round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
fmadd_ ⚠sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fmadd_ ⚠sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fmaddsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst. - _mm_
fmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst. _MM_FROUND_CUR_DIRECTION - _mm_
fmsub_ ⚠round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
fmsub_ ⚠round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fmsub_ ⚠round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
fmsub_ ⚠sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fmsubadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst. - _mm_
fmul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fmul_ ⚠round_ sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fmul_ ⚠sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
fnmadd_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst. - _mm_
fnmadd_ ⚠round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
fnmadd_ ⚠round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fnmadd_ ⚠round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
fnmadd_ ⚠sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fnmsub_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst. - _mm_
fnmsub_ ⚠round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
fnmsub_ ⚠round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fnmsub_ ⚠round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, subtract the lower element in c from the negated intermediate result, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
fnmsub_ ⚠sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
fpclass_ ⚠pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm_
fpclass_ ⚠ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm_
fpclass_ ⚠ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm_
fpclass_ ⚠sd_ mask Experimental Test the lower double-precision (64-bit) floating-point element in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm_
fpclass_ ⚠sh_ mask Experimental Test the lower half-precision (16-bit) floating-point element in a for special categories specified by imm8, and store the result in mask vector k. imm can be a combination of: - _mm_
fpclass_ ⚠ss_ mask Experimental Test the lower single-precision (32-bit) floating-point element in a for special categories specified by imm8, and store the results in mask vector k. imm can be a combination of: - _mm_
getexp_ ⚠pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
getexp_ ⚠ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm_
getexp_ ⚠ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst. This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
getexp_ ⚠round_ sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
getexp_ ⚠round_ sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
getexp_ ⚠round_ ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
getexp_ ⚠sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
getexp_ ⚠sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. - _mm_
getexp_ ⚠ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
getmant_ ⚠pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
getmant_ ⚠ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
getmant_ ⚠ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign. The mantissa is normalized to the interval specified by interv, which can take the following values: _MM_MANT_NORM_1_2 // interval [1, 2) _MM_MANT_NORM_p5_2 // interval [0.5, 2) _MM_MANT_NORM_p5_1 // interval [0.5, 1) _MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5) The sign is determined by sc which can take the following values: _MM_MANT_SIGN_src // sign = sign(src) _MM_MANT_SIGN_zero // sign = 0 _MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
getmant_ ⚠round_ sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
getmant_ ⚠round_ sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
getmant_ ⚠round_ ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
getmant_ ⚠sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
getmant_ ⚠sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
getmant_ ⚠ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
gf2p8affine_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
gf2p8affineinv_ ⚠epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
gf2p8mul_ ⚠epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm_
i32scatter_ ⚠epi32 Experimental Stores 4 32-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm_
i32scatter_ ⚠epi64 Experimental Stores 2 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm_
i32scatter_ ⚠pd Experimental Stores 2 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm_
i32scatter_ ⚠ps Experimental Stores 4 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale - _mm_
i64scatter_ ⚠epi32 Experimental Stores 2 32-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm_
i64scatter_ ⚠epi64 Experimental Stores 2 64-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm_
i64scatter_ ⚠pd Experimental Stores 2 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm_
i64scatter_ ⚠ps Experimental Stores 2 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale - _mm_
load_ ⚠epi32 Experimental Load 128-bits (composed of 4 packed 32-bit integers) from memory into dst. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
load_ ⚠epi64 Experimental Load 128-bits (composed of 2 packed 64-bit integers) from memory into dst. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
load_ ⚠ph Experimental Load 128-bits (composed of 8 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address must be aligned to 16 bytes or a general-protection exception may be generated. - _mm_
load_ ⚠sh Experimental Load a half-precision (16-bit) floating-point element from memory into the lower element of a new vector, and zero the upper elements - _mm_
loadu_ ⚠epi8 Experimental Load 128-bits (composed of 16 packed 8-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm_
loadu_ ⚠epi16 Experimental Load 128-bits (composed of 8 packed 16-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm_
loadu_ ⚠epi32 Experimental Load 128-bits (composed of 4 packed 32-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm_
loadu_ ⚠epi64 Experimental Load 128-bits (composed of 2 packed 64-bit integers) from memory into dst. mem_addr does not need to be aligned on any particular boundary. - _mm_
loadu_ ⚠ph Experimental Load 128-bits (composed of 8 packed half-precision (16-bit) floating-point elements) from memory into a new vector. The address does not need to be aligned to any particular boundary. - _mm_
lzcnt_ ⚠epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst. - _mm_
lzcnt_ ⚠epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst. - _mm_
madd52hi_ ⚠avx_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm_
madd52hi_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm_
madd52lo_ ⚠avx_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm_
madd52lo_ ⚠epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
. - _mm_
mask2_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask2_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm_
mask2_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm_
mask2_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm_
mask2_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set) - _mm_
mask2_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from idx when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask3_ ⚠fcmadd_ round_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (the element is copied from c when the corresponding mask bit is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask3_ ⚠fcmadd_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (the element is copied from c when the corresponding mask bit is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask3_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask3_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmadd_ round_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (elements are copied from c when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask3_ ⚠fmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst.\ - _mm_
mask3_ ⚠fmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst.\ - _mm_
mask3_ ⚠fmadd_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (elements are copied from c when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask3_ ⚠fmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst. - _mm_
mask3_ ⚠fmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmaddsub_ pd Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst.\ - _mm_
mask3_ ⚠fmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst.\ - _mm_
mask3_ ⚠fmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst. - _mm_
mask3_ ⚠fmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst.\ - _mm_
mask3_ ⚠fnmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fnmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst.\ - _mm_
mask3_ ⚠fnmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst. - _mm_
mask3_ ⚠fnmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fnmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from c when the corresponding mask bit is not set). - _mm_
mask3_ ⚠fnmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst.\ - _mm_
mask3_ ⚠fnmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fnmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst.\ - _mm_
mask3_ ⚠fnmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from c to the upper element of dst. - _mm_
mask3_ ⚠fnmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from c when the mask bit 0 is not set), and copy the upper 7 packed elements from c to the upper elements of dst. - _mm_
mask3_ ⚠fnmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from c to the upper elements of dst. - _mm_
mask_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set) - _mm_
mask_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠abs_ epi32 Experimental Compute the absolute value of packed signed 32-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠add_ round_ sd Experimental Add the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠add_ round_ sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
mask_ ⚠add_ round_ ss Experimental Add the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠add_ sd Experimental Add the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠add_ sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). - _mm_
mask_ ⚠add_ ss Experimental Add the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 16 bytes (4 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 16 bytes (2 elements) in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠and_ epi32 Experimental Performs element-by-element bitwise AND between packed 32-bit integer elements of a and b, storing the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠bitshuffle_ epi64_ mask Experimental Considers the inputb
as packed 64-bit integers andc
as packed 8-bit integers. Then groups 8 8-bit values fromc
as indices into the bits of the corresponding 64-bit integer. It then selects these bits and packs them into the output. - _mm_
mask_ ⚠blend_ epi8 Experimental Blend packed 8-bit integers from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ epi16 Experimental Blend packed 16-bit integers from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ epi32 Experimental Blend packed 32-bit integers from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ epi64 Experimental Blend packed 64-bit integers from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ pd Experimental Blend packed double-precision (64-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ ph Experimental Blend packed half-precision (16-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠blend_ ps Experimental Blend packed single-precision (32-bit) floating-point elements from a and b using control mask k, and store the results in dst. - _mm_
mask_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ pd_ mask Experimental Compare packed double-precision (64-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ ph_ mask Experimental Compare packed half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ ps_ mask Experimental Compare packed single-precision (32-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmp_ round_ sd_ mask Experimental Compare the lower double-precision (64-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1 (the element is zeroed out when mask bit 0 is not set).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠cmp_ round_ sh_ mask Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠cmp_ round_ ss_ mask Experimental Compare the lower single-precision (32-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1 (the element is zeroed out when mask bit 0 is not seti).
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠cmp_ sd_ mask Experimental Compare the lower double-precision (64-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1 (the element is zeroed out when mask bit 0 is not set). - _mm_
mask_ ⚠cmp_ sh_ mask Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1. - _mm_
mask_ ⚠cmp_ ss_ mask Experimental Compare the lower single-precision (32-bit) floating-point element in a and b based on the comparison operand specified by imm8, and store the result in mask vector k using zeromask k1 (the element is zeroed out when mask bit 0 is not set). - _mm_
mask_ ⚠cmpeq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epi64_ mask Experimental Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpeq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpge_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpgt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmple_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epi32_ mask Experimental Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmplt_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epi8_ mask Experimental Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epi16_ mask Experimental Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epi32_ mask Experimental Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epi64_ mask Experimental Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epu8_ mask Experimental Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epu16_ mask Experimental Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epu32_ mask Experimental Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmpneq_ epu64_ mask Experimental Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). - _mm_
mask_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠cmul_ round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠cmul_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm_
mask_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to dst, and pass through the remaining elements from src. - _mm_
mask_ ⚠compressstoreu_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠compressstoreu_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠compressstoreu_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠compressstoreu_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠compressstoreu_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠compressstoreu_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm_
mask_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using writemask k (elements are copied from src when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm_
mask_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠cvt_ roundsd_ sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using writemask k (the element if copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvt_ roundsd_ ss Experimental Convert the lower double-precision (64-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠cvt_ roundsh_ sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src to dst when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠cvt_ roundsh_ ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src to dst when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvt_ roundss_ sd Experimental Convert the lower single-precision (32-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠cvt_ roundss_ sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using writemask k (the element if copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 2 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi16_ storeu_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
mask_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi32_ storeu_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepi32_ storeu_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
mask_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtepi64_ storeu_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepi64_ storeu_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepi64_ storeu_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in the low 4 bytes of a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 2 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
mask_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
mask_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in single vector dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm_
mask_ ⚠cvtneps_ pbh Experimental Converts packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
mask_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠cvtsd_ sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using writemask k (the element if copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvtsd_ ss Experimental Convert the lower double-precision (64-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi16_ storeu_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi32_ storeu_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsepi32_ storeu_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtsepi64_ storeu_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsepi64_ storeu_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsepi64_ storeu_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtsh_ sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src to dst when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠cvtsh_ ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src to dst when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvtss_ sd Experimental Convert the lower single-precision (32-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠cvtss_ sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using writemask k (the element if copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi16_ storeu_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi32_ storeu_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtusepi32_ storeu_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtusepi64_ storeu_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed 8-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtusepi64_ storeu_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed 16-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtusepi64_ storeu_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed 32-bit integers with unsigned saturation, and store the active results (those with their respective bit set in writemask k) to unaligned memory at base_addr. - _mm_
mask_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). - _mm_
mask_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using writemask k (elements are copied from src to dst when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
mask_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm_
mask_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠div_ round_ sd Experimental Divide the lower double-precision (64-bit) floating-point element in a by the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠div_ round_ sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
mask_ ⚠div_ round_ ss Experimental Divide the lower single-precision (32-bit) floating-point element in a by the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠div_ sd Experimental Divide the lower double-precision (64-bit) floating-point element in a by the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠div_ sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). - _mm_
mask_ ⚠div_ ss Experimental Divide the lower single-precision (32-bit) floating-point element in a by the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Intel’s documentation - _mm_
mask_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fcmadd_ round_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (the element is copied from a when the corresponding mask bit is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fcmadd_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (the element is copied from a when the corresponding mask bit is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fcmul_ round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fcmul_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
mask_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm_
mask_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm_
mask_ ⚠fixupimm_ round_ sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠fixupimm_ round_ ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠fixupimm_ sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting. - _mm_
mask_ ⚠fixupimm_ ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting. - _mm_
mask_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmadd_ round_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (elements are copied from a when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠fmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠fmadd_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using writemask k (elements are copied from a when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠fmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠fmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠fmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠fmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fmul_ round_ sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fmul_ sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠fnmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fnmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠fnmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠fnmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fnmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (the element is copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠fnmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠fnmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fnmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠fnmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠fnmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using writemask k (the element is copied from a when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fnmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using writemask k (the element is copied from c when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠fpclass_ pd_ mask Experimental Test packed double-precision (64-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠fpclass_ ph_ mask Experimental Test packed half-precision (16-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠fpclass_ ps_ mask Experimental Test packed single-precision (32-bit) floating-point elements in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠fpclass_ sd_ mask Experimental Test the lower double-precision (64-bit) floating-point element in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠fpclass_ sh_ mask Experimental Test the lower half-precision (16-bit) floating-point element in a for special categories specified by imm8, and store the result in mask vector k using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠fpclass_ ss_ mask Experimental Test the lower single-precision (32-bit) floating-point element in a for special categories specified by imm8, and store the results in mask vector k using zeromask k1 (elements are zeroed out when the corresponding mask bit is not set). imm can be a combination of: - _mm_
mask_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
mask_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm_
mask_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
mask_ ⚠getexp_ round_ sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠getexp_ round_ sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
mask_ ⚠getexp_ round_ ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠getexp_ sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
mask_ ⚠getexp_ sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. - _mm_
mask_ ⚠getexp_ ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
mask_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
mask_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
mask_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
mask_ ⚠getmant_ round_ sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠getmant_ round_ sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
mask_ ⚠getmant_ round_ ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠getmant_ sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠getmant_ sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
mask_ ⚠getmant_ ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
mask_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
mask_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm_
mask_ ⚠i32scatter_ epi32 Experimental Stores 4 32-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i32scatter_ epi64 Experimental Stores 2 64-bit integer elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i32scatter_ pd Experimental Stores 2 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i32scatter_ ps Experimental Stores 4 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i64scatter_ epi32 Experimental Stores 2 32-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i64scatter_ epi64 Experimental Stores 2 64-bit integer elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i64scatter_ pd Experimental Stores 2 double-precision (64-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding mask bit is not set are not written to memory). - _mm_
mask_ ⚠i64scatter_ ps Experimental Stores 2 single-precision (32-bit) floating-point elements from a to memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements whose corresponding - _mm_
mask_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠load_ sd Experimental Load a double-precision (64-bit) floating-point element from memory into the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and set the upper element of dst to zero. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠load_ sh Experimental Load a half-precision (16-bit) floating-point element from memory into the lower element of a new vector using writemask k (the element is copied from src when mask bit 0 is not set), and zero the upper elements. - _mm_
mask_ ⚠load_ ss Experimental Load a single-precision (32-bit) floating-point element from memory into the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and set the upper 3 packed elements of dst to zero. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm_
mask_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are copied fromk
when the corresponding mask bit is not set). - _mm_
mask_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠max_ round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠max_ round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠max_ round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠max_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠max_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠max_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠min_ round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠min_ round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠min_ round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠min_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠min_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
mask_ ⚠min_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠move_ sd Experimental Move the lower double-precision (64-bit) floating-point element from b to the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠move_ sh Experimental Move the lower half-precision (16-bit) floating-point element from b to the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠move_ ss Experimental Move the lower single-precision (32-bit) floating-point element from b to the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using writemask k (the element is copied from src when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mul_ round_ sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠mul_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠mul_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
mask_ ⚠mul_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠mul_ sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mask_ ⚠mul_ sd Experimental Multiply the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠mul_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). - _mm_
mask_ ⚠mul_ ss Experimental Multiply the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using writemaskk
(elements are copied fromsrc
if the corresponding bit is not set). - _mm_
mask_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm_
mask_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm_
mask_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm_
mask_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm_
mask_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
mask_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
mask_ ⚠range_ round_ sd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠range_ round_ ss Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
mask_ ⚠range_ sd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
mask_ ⚠range_ ss Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
mask_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rcp14_ sd Experimental Compute the approximate reciprocal of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rcp14_ ss Experimental Compute the approximate reciprocal of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using writemaskk
(elements are copied fromsrc
when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
mask_ ⚠rcp_ sh Experimental Compute the approximate reciprocal of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
mask_ ⚠reduce_ add_ epi8 Experimental Reduce the packed 8-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm_
mask_ ⚠reduce_ add_ epi16 Experimental Reduce the packed 16-bit integers in a by addition using mask k. Returns the sum of all active elements in a. - _mm_
mask_ ⚠reduce_ and_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise AND using mask k. Returns the bitwise AND of all active elements in a. - _mm_
mask_ ⚠reduce_ and_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise AND using mask k. Returns the bitwise AND of all active elements in a. - _mm_
mask_ ⚠reduce_ max_ epi8 Experimental Reduce the packed 8-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm_
mask_ ⚠reduce_ max_ epi16 Experimental Reduce the packed 16-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm_
mask_ ⚠reduce_ max_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm_
mask_ ⚠reduce_ max_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by maximum using mask k. Returns the maximum of all active elements in a. - _mm_
mask_ ⚠reduce_ min_ epi8 Experimental Reduce the packed 8-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm_
mask_ ⚠reduce_ min_ epi16 Experimental Reduce the packed 16-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm_
mask_ ⚠reduce_ min_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm_
mask_ ⚠reduce_ min_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by minimum using mask k. Returns the minimum of all active elements in a. - _mm_
mask_ ⚠reduce_ mul_ epi8 Experimental Reduce the packed 8-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm_
mask_ ⚠reduce_ mul_ epi16 Experimental Reduce the packed 16-bit integers in a by multiplication using mask k. Returns the product of all active elements in a. - _mm_
mask_ ⚠reduce_ or_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm_
mask_ ⚠reduce_ or_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise OR using mask k. Returns the bitwise OR of all active elements in a. - _mm_
mask_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using writemask k (elements are copied from src to dst if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠reduce_ round_ sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠reduce_ round_ sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠reduce_ round_ ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠reduce_ sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠reduce_ sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠reduce_ ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
mask_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠roundscale_ round_ sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
mask_ ⚠roundscale_ round_ sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠roundscale_ round_ ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
mask_ ⚠roundscale_ sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠roundscale_ sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠roundscale_ ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
mask_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rsqrt14_ sd Experimental Compute the approximate reciprocal square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rsqrt14_ ss Experimental Compute the approximate reciprocal square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
mask_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
mask_ ⚠rsqrt_ sh Experimental Compute the approximate reciprocal square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
mask_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠scalef_ round_ sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠scalef_ round_ sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠scalef_ round_ ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠scalef_ sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠scalef_ sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠scalef_ ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠set1_ epi16 Experimental Broadcast 16-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using writemask k (elements are copied from src“ when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using writemask k (elements are copied from a when the corresponding mask bit is not set). - _mm_
mask_ ⚠shuffle_ epi8 Experimental Shuffle 8-bit integers in a within 128-bit lanes using the control in the corresponding 8-bit element of b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a using the control in imm8, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sqrt_ round_ sd Experimental Compute the square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠sqrt_ round_ sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
mask_ ⚠sqrt_ round_ ss Experimental Compute the square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠sqrt_ sd Experimental Compute the square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠sqrt_ sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠sqrt_ ss Experimental Compute the square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠store_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠store_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠store_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠store_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠store_ sd Experimental Store a double-precision (64-bit) floating-point element from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠store_ sh Experimental Store the lower half-precision (16-bit) floating-point element from a into memory using writemask k - _mm_
mask_ ⚠store_ ss Experimental Store a single-precision (32-bit) floating-point element from a into memory using writemask k. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
mask_ ⚠storeu_ epi8 Experimental Store packed 8-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠storeu_ epi16 Experimental Store packed 16-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠storeu_ epi32 Experimental Store packed 32-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠storeu_ epi64 Experimental Store packed 64-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠storeu_ pd Experimental Store packed double-precision (64-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠storeu_ ps Experimental Store packed single-precision (32-bit) floating-point elements from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary. - _mm_
mask_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠sub_ round_ sd Experimental Subtract the lower double-precision (64-bit) floating-point element in b from the lower double-precision (64-bit) floating-point element in a, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
mask_ ⚠sub_ round_ sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
mask_ ⚠sub_ round_ ss Experimental Subtract the lower single-precision (32-bit) floating-point element in b from the lower single-precision (32-bit) floating-point element in a, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mask_ ⚠sub_ sd Experimental Subtract the lower double-precision (64-bit) floating-point element in b from the lower double-precision (64-bit) floating-point element in a, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
mask_ ⚠sub_ sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using writemask k (the element is copied from src when mask bit 0 is not set). - _mm_
mask_ ⚠sub_ ss Experimental Subtract the lower single-precision (32-bit) floating-point element in b from the lower single-precision (32-bit) floating-point element in a, store the result in the lower element of dst using writemask k (the element is copied from src when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
mask_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 32-bit granularity (32-bit elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from src, a, and b are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using writemask k at 64-bit granularity (64-bit elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠test_ epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm_
mask_ ⚠test_ epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm_
mask_ ⚠test_ epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm_
mask_ ⚠test_ epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is non-zero. - _mm_
mask_ ⚠testn_ epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm_
mask_ ⚠testn_ epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm_
mask_ ⚠testn_ epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm_
mask_ ⚠testn_ epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k (subject to writemask k) if the intermediate value is zero. - _mm_
mask_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mask_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
mask_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using writemask k (elements are copied from src if the corresponding bit is not set). - _mm_
maskz_ ⚠abs_ epi8 Experimental Compute the absolute value of packed signed 8-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠abs_ epi16 Experimental Compute the absolute value of packed signed 16-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠abs_ epi32 Experimental Compute the absolute value of packed signed 32-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠abs_ epi64 Experimental Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ epi8 Experimental Add packed 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ epi16 Experimental Add packed 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ epi32 Experimental Add packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ epi64 Experimental Add packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ pd Experimental Add packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ ph Experimental Add packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ ps Experimental Add packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠add_ round_ sd Experimental Add the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠add_ round_ sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
maskz_ ⚠add_ round_ ss Experimental Add the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠add_ sd Experimental Add the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠add_ sh Experimental Add the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). - _mm_
maskz_ ⚠add_ ss Experimental Add the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠adds_ epi8 Experimental Add packed signed 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠adds_ epi16 Experimental Add packed signed 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠adds_ epu8 Experimental Add packed unsigned 8-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠adds_ epu16 Experimental Add packed unsigned 16-bit integers in a and b using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠alignr_ epi8 Experimental Concatenate pairs of 16-byte blocks in a and b into a 32-byte temporary result, shift the result right by imm8 bytes, and store the low 16 bytes in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠alignr_ epi32 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 32-bit elements, and store the low 16 bytes (4 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠alignr_ epi64 Experimental Concatenate a and b into a 32-byte immediate result, shift the result right by imm8 64-bit elements, and store the low 16 bytes (2 elements) in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠and_ epi32 Experimental Compute the bitwise AND of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠and_ epi64 Experimental Compute the bitwise AND of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠and_ pd Experimental Compute the bitwise AND of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠and_ ps Experimental Compute the bitwise AND of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠andnot_ epi32 Experimental Compute the bitwise NOT of packed 32-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠andnot_ epi64 Experimental Compute the bitwise NOT of packed 64-bit integers in a and then AND with b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠andnot_ pd Experimental Compute the bitwise NOT of packed double-precision (64-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠andnot_ ps Experimental Compute the bitwise NOT of packed single-precision (32-bit) floating point numbers in a and then bitwise AND with b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠avg_ epu8 Experimental Average packed unsigned 8-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠avg_ epu16 Experimental Average packed unsigned 16-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠broadcast_ i32x2 Experimental Broadcasts the lower 2 packed 32-bit integers from a to all elements of dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠broadcastb_ epi8 Experimental Broadcast the low packed 8-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠broadcastd_ epi32 Experimental Broadcast the low packed 32-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠broadcastq_ epi64 Experimental Broadcast the low packed 64-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠broadcastss_ ps Experimental Broadcast the low single-precision (32-bit) floating-point element from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠broadcastw_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠cmul_ round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠cmul_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, - _mm_
maskz_ ⚠compress_ epi8 Experimental Contiguously store the active 8-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠compress_ epi16 Experimental Contiguously store the active 16-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠compress_ epi32 Experimental Contiguously store the active 32-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠compress_ epi64 Experimental Contiguously store the active 64-bit integers in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠compress_ pd Experimental Contiguously store the active double-precision (64-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠compress_ ps Experimental Contiguously store the active single-precision (32-bit) floating-point elements in a (those with their respective bit set in zeromask k) to dst, and set the remaining elements to zero. - _mm_
maskz_ ⚠conflict_ epi32 Experimental Test each 32-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm_
maskz_ ⚠conflict_ epi64 Experimental Test each 64-bit element of a for equality with all other elements in a closer to the least significant bit using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Each element’s comparison forms a zero extended bit vector in dst. - _mm_
maskz_ ⚠conj_ pch Experimental Compute the complex conjugates of complex numbers in a, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠cvt_ roundps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠cvt_ roundsd_ sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvt_ roundsd_ ss Experimental Convert the lower double-precision (64-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the rounding[3:0] parameter, which can be one of:
(_MM_FROUND_TO_NEAREST_INT |_MM_FROUND_NO_EXC) // round to nearest, and suppress exceptions
(_MM_FROUND_TO_NEG_INF |_MM_FROUND_NO_EXC) // round down, and suppress exceptions
(_MM_FROUND_TO_POS_INF |_MM_FROUND_NO_EXC) // round up, and suppress exceptions
(_MM_FROUND_TO_ZERO |_MM_FROUND_NO_EXC) // truncate, and suppress exceptions
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠cvt_ roundsh_ sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠cvt_ roundsh_ ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvt_ roundss_ sd Experimental Convert the lower single-precision (32-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠cvt_ roundss_ sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvtepi8_ epi16 Experimental Sign extend packed 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi8_ epi32 Experimental Sign extend packed 8-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi8_ epi64 Experimental Sign extend packed 8-bit integers in the low 2 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi16_ epi8 Experimental Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi16_ epi32 Experimental Sign extend packed 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi16_ epi64 Experimental Sign extend packed 16-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi16_ ph Experimental Convert packed signed 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi32_ epi8 Experimental Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi32_ epi16 Experimental Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi32_ epi64 Experimental Sign extend packed 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi32_ pd Experimental Convert packed signed 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi32_ ph Experimental Convert packed signed 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
maskz_ ⚠cvtepi32_ ps Experimental Convert packed signed 32-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi64_ epi8 Experimental Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi64_ epi16 Experimental Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi64_ epi32 Experimental Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepi64_ pd Experimental Convert packed signed 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtepi64_ ph Experimental Convert packed signed 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
maskz_ ⚠cvtepi64_ ps Experimental Convert packed signed 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtepu8_ epi16 Experimental Zero extend packed unsigned 8-bit integers in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu8_ epi32 Experimental Zero extend packed unsigned 8-bit integers in th elow 4 bytes of a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu8_ epi64 Experimental Zero extend packed unsigned 8-bit integers in the low 2 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu16_ epi32 Experimental Zero extend packed unsigned 16-bit integers in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu16_ epi64 Experimental Zero extend packed unsigned 16-bit integers in the low 4 bytes of a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu16_ ph Experimental Convert packed unsigned 16-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu32_ epi64 Experimental Zero extend packed unsigned 32-bit integers in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu32_ pd Experimental Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtepu32_ ph Experimental Convert packed unsigned 32-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
maskz_ ⚠cvtepu64_ pd Experimental Convert packed unsigned 64-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtepu64_ ph Experimental Convert packed unsigned 64-bit integers in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
maskz_ ⚠cvtepu64_ ps Experimental Convert packed unsigned 64-bit integers in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtne2ps_ pbh Experimental Convert packed single-precision (32-bit) floating-point elements in two vectors a and b to packed BF16 (16-bit) floating-point elements, and store the results in single vector dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm_
maskz_ ⚠cvtneps_ pbh Experimental Converts packed single-precision (32-bit) floating-point elements in a to packed BF16 (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtpbh_ ps Experimental Converts packed BF16 (16-bit) floating-point elements in a to single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtpd_ ph Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 96 bits of dst are zeroed out. - _mm_
maskz_ ⚠cvtpd_ ps Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ pd Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed double-precision (64-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtps_ epu32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers, and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠cvtsd_ sh Experimental Convert the lower double-precision (64-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvtsd_ ss Experimental Convert the lower double-precision (64-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvtsepi16_ epi8 Experimental Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtsepi32_ epi8 Experimental Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtsepi32_ epi16 Experimental Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst. - _mm_
maskz_ ⚠cvtsepi64_ epi8 Experimental Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtsepi64_ epi16 Experimental Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtsepi64_ epi32 Experimental Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtsh_ sd Experimental Convert the lower half-precision (16-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠cvtsh_ ss Experimental Convert the lower half-precision (16-bit) floating-point element in b to a single-precision (32-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvtss_ sd Experimental Convert the lower single-precision (32-bit) floating-point element in b to a double-precision (64-bit) floating-point element, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠cvtss_ sh Experimental Convert the lower single-precision (32-bit) floating-point element in b to a half-precision (16-bit) floating-point elements, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠cvttpd_ epi32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttpd_ epi64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvttpd_ epu32 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttpd_ epu64 Experimental Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvttph_ epi16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttph_ epi32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttph_ epi64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttph_ epu16 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed unsigned 16-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttph_ epu32 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 32-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttph_ epu64 Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed 64-bit unsigned integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttps_ epi32 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttps_ epi64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed signed 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvttps_ epu32 Experimental Convert packed double-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers with truncation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvttps_ epu64 Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 64-bit integers with truncation, and store the result in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠cvtusepi16_ epi8 Experimental Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtusepi32_ epi8 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtusepi32_ epi16 Experimental Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtusepi64_ epi8 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtusepi64_ epi16 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtusepi64_ epi32 Experimental Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtxph_ ps Experimental Convert packed half-precision (16-bit) floating-point elements in a to packed single-precision (32-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠cvtxps_ ph Experimental Convert packed single-precision (32-bit) floating-point elements in a to packed half-precision (16-bit) floating-point elements, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The upper 64 bits of dst are zeroed out. - _mm_
maskz_ ⚠dbsad_ epu8 Experimental Compute the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers in a compared to those in b, and store the 16-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Four SADs are performed on four 8-bit quadruplets for each 64-bit lane. The first two SADs use the lower 8-bit quadruplet of the lane from a, and the last two SADs use the uppper 8-bit quadruplet of the lane from a. Quadruplets from b are selected from within 128-bit lanes according to the control in imm8, and each SAD in each 64-bit lane uses the selected quadruplet at 8-bit offsets. - _mm_
maskz_ ⚠div_ pd Experimental Divide packed double-precision (64-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠div_ ph Experimental Divide packed half-precision (16-bit) floating-point elements in a by b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠div_ ps Experimental Divide packed single-precision (32-bit) floating-point elements in a by packed elements in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠div_ round_ sd Experimental Divide the lower double-precision (64-bit) floating-point element in a by the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠div_ round_ sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
maskz_ ⚠div_ round_ ss Experimental Divide the lower single-precision (32-bit) floating-point element in a by the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠div_ sd Experimental Divide the lower double-precision (64-bit) floating-point element in a by the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠div_ sh Experimental Divide the lower half-precision (16-bit) floating-point elements in a by b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). - _mm_
maskz_ ⚠div_ ss Experimental Divide the lower single-precision (32-bit) floating-point element in a by the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠dpbf16_ ps Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in src, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Intel’s documentation - _mm_
maskz_ ⚠dpbusd_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠dpbusds_ epi32 Experimental Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate signed 16-bit results. Sum these 4 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠dpwssd_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠dpwssds_ epi32 Experimental Multiply groups of 2 adjacent pairs of signed 16-bit integers in a with corresponding 16-bit integers in b, producing 2 intermediate signed 32-bit results. Sum these 2 results with the corresponding 32-bit integer in src using signed saturation, and store the packed 32-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ epi8 Experimental Load contiguous active 8-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ epi16 Experimental Load contiguous active 16-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ epi32 Experimental Load contiguous active 32-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ epi64 Experimental Load contiguous active 64-bit integers from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expand_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from a (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ epi8 Experimental Load contiguous active 8-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ epi16 Experimental Load contiguous active 16-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ epi32 Experimental Load contiguous active 32-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ epi64 Experimental Load contiguous active 64-bit integers from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ pd Experimental Load contiguous active double-precision (64-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠expandloadu_ ps Experimental Load contiguous active single-precision (32-bit) floating-point elements from unaligned memory at mem_addr (those with their respective bit set in mask k), and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fcmadd_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fcmadd_ round_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c using zeromask k (the element is zeroed out when the corresponding mask bit is not set), and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fcmadd_ sch Experimental Multiply the lower complex number in a by the complex conjugate of the lower complex number in b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fcmul_ pch Experimental Multiply packed complex numbers in a by the complex conjugates of packed complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fcmul_ round_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fcmul_ sch Experimental Multiply the lower complex numbers in a by the complex conjugates of the lower complex numbers in b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
, or the complex conjugateconjugate = vec.fp16[0] - i * vec.fp16[1]
. - _mm_
maskz_ ⚠fixupimm_ pd Experimental Fix up packed double-precision (64-bit) floating-point elements in a and b using packed 64-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm_
maskz_ ⚠fixupimm_ ps Experimental Fix up packed single-precision (32-bit) floating-point elements in a and b using packed 32-bit integers in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). imm8 is used to set the required flags reporting. - _mm_
maskz_ ⚠fixupimm_ round_ sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠fixupimm_ round_ ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠fixupimm_ sd Experimental Fix up the lower double-precision (64-bit) floating-point elements in a and b using the lower 64-bit integer in c, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. imm8 is used to set the required flags reporting. - _mm_
maskz_ ⚠fixupimm_ ss Experimental Fix up the lower single-precision (32-bit) floating-point elements in a and b using the lower 32-bit integer in c, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. imm8 is used to set the required flags reporting. - _mm_
maskz_ ⚠fmadd_ pch Experimental Multiply packed complex numbers in a and b, accumulate to the corresponding complex numbers in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmadd_ round_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using zeromask k (elements are zeroed out when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠fmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠fmadd_ sch Experimental Multiply the lower complex numbers in a and b, accumulate to the lower complex number in c, and store the result in the lower elements of dst using zeromask k (elements are zeroed out when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠fmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmaddsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmaddsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmaddsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠fmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠fmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠fmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract packed elements in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fmsubadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, alternatively add and subtract packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsubadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c to/from the intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmsubadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, alternatively subtract and add packed elements in c from/to the intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fmul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fmul_ round_ sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fmul_ sch Experimental Multiply the lower complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠fnmadd_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmadd_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract the intermediate result from packed elements in c, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmadd_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, add the negated intermediate result to packed elements in c, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmadd_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠fnmadd_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fnmadd_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠fnmadd_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠fnmadd_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fnmadd_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and add the negated intermediate result to the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fnmsub_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmsub_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (the element is zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmsub_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, subtract packed elements in c from the negated intermediate result, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠fnmsub_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠fnmsub_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fnmsub_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠fnmsub_ sd Experimental Multiply the lower double-precision (64-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠fnmsub_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, and subtract the intermediate result from the lower element in c. Store the result in the lower element of dst using zeromask k (the element is zeroed out when the mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠fnmsub_ ss Experimental Multiply the lower single-precision (32-bit) floating-point elements in a and b, and subtract the lower element in c from the negated intermediate result. Store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠getexp_ pd Experimental Convert the exponent of each packed double-precision (64-bit) floating-point element in a to a double-precision (64-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
maskz_ ⚠getexp_ ph Experimental Convert the exponent of each packed half-precision (16-bit) floating-point element in a to a half-precision (16-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculatesfloor(log2(x))
for each element. - _mm_
maskz_ ⚠getexp_ ps Experimental Convert the exponent of each packed single-precision (32-bit) floating-point element in a to a single-precision (32-bit) floating-point number representing the integer exponent, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates floor(log2(x)) for each element. - _mm_
maskz_ ⚠getexp_ round_ sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠getexp_ round_ sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
maskz_ ⚠getexp_ round_ ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠getexp_ sd Experimental Convert the exponent of the lower double-precision (64-bit) floating-point element in b to a double-precision (64-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
maskz_ ⚠getexp_ sh Experimental Convert the exponent of the lower half-precision (16-bit) floating-point element in b to a half-precision (16-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculatesfloor(log2(x))
for the lower element. - _mm_
maskz_ ⚠getexp_ ss Experimental Convert the exponent of the lower single-precision (32-bit) floating-point element in b to a single-precision (32-bit) floating-point number representing the integer exponent, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates floor(log2(x)) for the lower element. - _mm_
maskz_ ⚠getmant_ pd Experimental Normalize the mantissas of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
maskz_ ⚠getmant_ ph Experimental Normalize the mantissas of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
maskz_ ⚠getmant_ ps Experimental Normalize the mantissas of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1 - _mm_
maskz_ ⚠getmant_ round_ sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠getmant_ round_ sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter - _mm_
maskz_ ⚠getmant_ round_ ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠getmant_ sd Experimental Normalize the mantissas of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠getmant_ sh Experimental Normalize the mantissas of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. This intrinsic essentially calculates±(2^k)*|x.significand|
, where k depends on the interval range defined by norm and the sign depends on sign and the source sign. - _mm_
maskz_ ⚠getmant_ ss Experimental Normalize the mantissas of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. This intrinsic essentially calculates ±(2^k)*|x.significand|, where k depends on the interval range defined by interv and the sign depends on sc and the source sign.
The mantissa is normalized to the interval specified by interv, which can take the following values:
_MM_MANT_NORM_1_2 // interval [1, 2)
_MM_MANT_NORM_p5_2 // interval [0.5, 2)
_MM_MANT_NORM_p5_1 // interval [0.5, 1)
_MM_MANT_NORM_p75_1p5 // interval [0.75, 1.5)
The sign is determined by sc which can take the following values:
_MM_MANT_SIGN_src // sign = sign(src)
_MM_MANT_SIGN_zero // sign = 0
_MM_MANT_SIGN_nan // dst = NaN if sign(src) = 1
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠gf2p8affine_ epi64_ epi8 Experimental Performs an affine transformation on the packed bytes in x. That is computes a*x+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
maskz_ ⚠gf2p8affineinv_ epi64_ epi8 Experimental Performs an affine transformation on the inverted packed bytes in x. That is computes a*inv(x)+b over the Galois Field 2^8 for each packed byte with a being a 8x8 bit matrix and b being a constant 8-bit immediate value. The inverse of a byte is defined with respect to the reduction polynomial x^8+x^4+x^3+x+1. The inverse of 0 is 0. Each pack of 8 bytes in x is paired with the 64-bit word at the same position in a. - _mm_
maskz_ ⚠gf2p8mul_ epi8 Experimental Performs a multiplication in GF(2^8) on the packed bytes. The field is in polynomial representation with the reduction polynomial x^8 + x^4 + x^3 + x + 1. - _mm_
maskz_ ⚠load_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠load_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠load_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠load_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠load_ sd Experimental Load a double-precision (64-bit) floating-point element from memory into the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and set the upper element of dst to zero. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠load_ sh Experimental Load a half-precision (16-bit) floating-point element from memory into the lower element of a new vector using zeromask k (the element is zeroed out when mask bit 0 is not set), and zero the upper elements. - _mm_
maskz_ ⚠load_ ss Experimental Load a single-precision (32-bit) floating-point element from memory into the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and set the upper 3 packed elements of dst to zero. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
maskz_ ⚠loadu_ epi8 Experimental Load packed 8-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠loadu_ epi16 Experimental Load packed 16-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠loadu_ epi32 Experimental Load packed 32-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠loadu_ epi64 Experimental Load packed 64-bit integers from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠loadu_ pd Experimental Load packed double-precision (64-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠loadu_ ps Experimental Load packed single-precision (32-bit) floating-point elements from memory into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). mem_addr does not need to be aligned on any particular boundary. - _mm_
maskz_ ⚠lzcnt_ epi32 Experimental Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠lzcnt_ epi64 Experimental Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠madd52hi_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the high 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠madd52lo_ epu64 Experimental Multiply packed unsigned 52-bit integers in each 64-bit element ofb
andc
to form a 104-bit intermediate result. Add the low 52-bit unsigned integer from the intermediate result with the corresponding unsigned 64-bit integer ina
, and store the results indst
using writemaskk
(elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠madd_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers, and pack the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠maddubs_ epi16 Experimental Multiply packed unsigned 8-bit integers in a by packed signed 8-bit integers in b, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers, and pack the saturated results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠max_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed maximum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠max_ round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠max_ round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠max_ round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠max_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠max_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠max_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the maximum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠min_ epi8 Experimental Compare packed signed 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epi16 Experimental Compare packed signed 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epi32 Experimental Compare packed signed 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epu8 Experimental Compare packed unsigned 8-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epu16 Experimental Compare packed unsigned 16-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epu32 Experimental Compare packed unsigned 32-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ pd Experimental Compare packed double-precision (64-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠min_ ps Experimental Compare packed single-precision (32-bit) floating-point elements in a and b, and store packed minimum values in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠min_ round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠min_ round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠min_ round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠min_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠min_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
maskz_ ⚠min_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the minimum value in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠mov_ epi8 Experimental Move packed 8-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mov_ epi16 Experimental Move packed 16-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mov_ epi32 Experimental Move packed 32-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mov_ epi64 Experimental Move packed 64-bit integers from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mov_ pd Experimental Move packed double-precision (64-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mov_ ps Experimental Move packed single-precision (32-bit) floating-point elements from a into dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠move_ sd Experimental Move the lower double-precision (64-bit) floating-point element from b to the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠move_ sh Experimental Move the lower half-precision (16-bit) floating-point element from b to the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠move_ ss Experimental Move the lower single-precision (32-bit) floating-point element from b to the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠movedup_ pd Experimental Duplicate even-indexed double-precision (64-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠movehdup_ ps Experimental Duplicate odd-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠moveldup_ ps Experimental Duplicate even-indexed single-precision (32-bit) floating-point elements from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ epi32 Experimental Multiply the low signed 32-bit integers from each packed 64-bit element in a and b, and store the signed 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ epu32 Experimental Multiply the low unsigned 32-bit integers from each packed 64-bit element in a and b, and store the unsigned 64-bit results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ pch Experimental Multiply packed complex numbers in a and b, and store the results in dst using zeromask k (the element is zeroed out when corresponding mask bit is not set). Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠mul_ pd Experimental Multiply packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ ps Experimental Multiply packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mul_ round_ sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠mul_ round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠mul_ round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
maskz_ ⚠mul_ round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠mul_ sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
maskz_ ⚠mul_ sd Experimental Multiply the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠mul_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). - _mm_
maskz_ ⚠mul_ ss Experimental Multiply the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠mulhi_ epi16 Experimental Multiply the packed signed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mulhi_ epu16 Experimental Multiply the packed unsigned 16-bit integers in a and b, producing intermediate 32-bit integers, and store the high 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mulhrs_ epi16 Experimental Multiply packed signed 16-bit integers in a and b, producing intermediate signed 32-bit integers. Truncate each intermediate integer to the 18 most significant bits, round by adding 1, and store bits [16:1] to dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mullo_ epi16 Experimental Multiply the packed 16-bit integers in a and b, producing intermediate 32-bit integers, and store the low 16 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mullo_ epi32 Experimental Multiply the packed 32-bit integers in a and b, producing intermediate 64-bit integers, and store the low 32 bits of the intermediate integers in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠mullo_ epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
using zeromaskk
(elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠multishift_ epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠or_ epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠or_ epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠or_ pd Experimental Compute the bitwise OR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠or_ ps Experimental Compute the bitwise OR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠packs_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠packs_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using signed saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠packus_ epi16 Experimental Convert packed signed 16-bit integers from a and b to packed 8-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠packus_ epi32 Experimental Convert packed signed 32-bit integers from a and b to packed 16-bit integers using unsigned saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permute_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permute_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutevar_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutevar_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a within 128-bit lanes using the control in b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutex2var_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutexvar_ epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠permutexvar_ epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠popcnt_ epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm_
maskz_ ⚠popcnt_ epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm_
maskz_ ⚠popcnt_ epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm_
maskz_ ⚠popcnt_ epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm_
maskz_ ⚠range_ pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
maskz_ ⚠range_ ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
maskz_ ⚠range_ round_ sd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠range_ round_ ss Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
maskz_ ⚠range_ sd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
maskz_ ⚠range_ ss Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
maskz_ ⚠rcp14_ pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rcp14_ ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rcp14_ sd Experimental Compute the approximate reciprocal of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rcp14_ ss Experimental Compute the approximate reciprocal of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rcp_ ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
using zeromaskk
(elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
maskz_ ⚠rcp_ sh Experimental Compute the approximate reciprocal of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
maskz_ ⚠reduce_ pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠reduce_ ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠reduce_ ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out if the corresponding mask bit is not set). Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠reduce_ round_ sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠reduce_ round_ sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠reduce_ round_ ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠reduce_ sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠reduce_ sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠reduce_ ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
maskz_ ⚠rol_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠rol_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠rolv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠rolv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠ror_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠ror_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠rorv_ epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠rorv_ epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠roundscale_ pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠roundscale_ ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠roundscale_ ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set).
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠roundscale_ round_ sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
maskz_ ⚠roundscale_ round_ sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠roundscale_ round_ ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
maskz_ ⚠roundscale_ sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠roundscale_ sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠roundscale_ ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
maskz_ ⚠rsqrt14_ pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rsqrt14_ ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rsqrt14_ sd Experimental Compute the approximate reciprocal square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rsqrt14_ ss Experimental Compute the approximate reciprocal square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
maskz_ ⚠rsqrt_ ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
maskz_ ⚠rsqrt_ sh Experimental Compute the approximate reciprocal square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
maskz_ ⚠scalef_ pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠scalef_ ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠scalef_ ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠scalef_ round_ sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠scalef_ round_ sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠scalef_ round_ ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠scalef_ sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠scalef_ sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠scalef_ ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠set1_ epi8 Experimental Broadcast 8-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠set1_ epi16 Experimental Broadcast the low packed 16-bit integer from a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠set1_ epi32 Experimental Broadcast 32-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠set1_ epi64 Experimental Broadcast 64-bit integer a to all elements of dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldi_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldi_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldi_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldv_ epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldv_ epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shldv_ epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdi_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdi_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdi_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdv_ epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdv_ epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shrdv_ epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shuffle_ epi8 Experimental Shuffle packed 8-bit integers in a according to shuffle control mask in the corresponding 8-bit element of b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shuffle_ epi32 Experimental Shuffle 32-bit integers in a within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shuffle_ pd Experimental Shuffle double-precision (64-bit) floating-point elements within 128-bit lanes using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shuffle_ ps Experimental Shuffle single-precision (32-bit) floating-point elements in a using the control in imm8, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shufflehi_ epi16 Experimental Shuffle 16-bit integers in the high 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the high 64 bits of 128-bit lanes of dst, with the low 64 bits of 128-bit lanes being copied from a to dst, using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠shufflelo_ epi16 Experimental Shuffle 16-bit integers in the low 64 bits of 128-bit lanes of a using the control in imm8. Store the results in the low 64 bits of 128-bit lanes of dst, with the high 64 bits of 128-bit lanes being copied from a to dst, using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sll_ epi16 Experimental Shift packed 16-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sll_ epi32 Experimental Shift packed 32-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sll_ epi64 Experimental Shift packed 64-bit integers in a left by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠slli_ epi16 Experimental Shift packed 16-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠slli_ epi32 Experimental Shift packed 32-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠slli_ epi64 Experimental Shift packed 64-bit integers in a left by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sllv_ epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sllv_ epi32 Experimental Shift packed 32-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sllv_ epi64 Experimental Shift packed 64-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sqrt_ pd Experimental Compute the square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sqrt_ ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sqrt_ ps Experimental Compute the square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sqrt_ round_ sd Experimental Compute the square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠sqrt_ round_ sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
maskz_ ⚠sqrt_ round_ ss Experimental Compute the square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠sqrt_ sd Experimental Compute the square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠sqrt_ sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠sqrt_ ss Experimental Compute the square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠sra_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sra_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sra_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srai_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srai_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srai_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srav_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srav_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srav_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srl_ epi16 Experimental Shift packed 16-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srl_ epi32 Experimental Shift packed 32-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srl_ epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srli_ epi16 Experimental Shift packed 16-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srli_ epi32 Experimental Shift packed 32-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srli_ epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srlv_ epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srlv_ epi32 Experimental Shift packed 32-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠srlv_ epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ epi8 Experimental Subtract packed 8-bit integers in b from packed 8-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ epi16 Experimental Subtract packed 16-bit integers in b from packed 16-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ epi32 Experimental Subtract packed 32-bit integers in b from packed 32-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ epi64 Experimental Subtract packed 64-bit integers in b from packed 64-bit integers in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ pd Experimental Subtract packed double-precision (64-bit) floating-point elements in b from packed double-precision (64-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ ps Experimental Subtract packed single-precision (32-bit) floating-point elements in b from packed single-precision (32-bit) floating-point elements in a, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠sub_ round_ sd Experimental Subtract the lower double-precision (64-bit) floating-point element in b from the lower double-precision (64-bit) floating-point element in a, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst.\ - _mm_
maskz_ ⚠sub_ round_ sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). Rounding is done according to the rounding parameter, which can be one of: - _mm_
maskz_ ⚠sub_ round_ ss Experimental Subtract the lower single-precision (32-bit) floating-point element in b from the lower single-precision (32-bit) floating-point element in a, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
maskz_ ⚠sub_ sd Experimental Subtract the lower double-precision (64-bit) floating-point element in b from the lower double-precision (64-bit) floating-point element in a, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper element from a to the upper element of dst. - _mm_
maskz_ ⚠sub_ sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst using zeromask k (the element is zeroed out when mask bit 0 is not set). - _mm_
maskz_ ⚠sub_ ss Experimental Subtract the lower single-precision (32-bit) floating-point element in b from the lower single-precision (32-bit) floating-point element in a, store the result in the lower element of dst using zeromask k (the element is zeroed out when mask bit 0 is not set), and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
maskz_ ⚠subs_ epi8 Experimental Subtract packed signed 8-bit integers in b from packed 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠subs_ epi16 Experimental Subtract packed signed 16-bit integers in b from packed 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠subs_ epu8 Experimental Subtract packed unsigned 8-bit integers in b from packed unsigned 8-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠subs_ epu16 Experimental Subtract packed unsigned 16-bit integers in b from packed unsigned 16-bit integers in a using saturation, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠ternarylogic_ epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 32-bit granularity (32-bit elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠ternarylogic_ epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst using zeromask k at 64-bit granularity (64-bit elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ epi8 Experimental Unpack and interleave 8-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ epi16 Experimental Unpack and interleave 16-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ epi32 Experimental Unpack and interleave 32-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ epi64 Experimental Unpack and interleave 64-bit integers from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpackhi_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ epi8 Experimental Unpack and interleave 8-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ epi16 Experimental Unpack and interleave 16-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ epi32 Experimental Unpack and interleave 32-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ epi64 Experimental Unpack and interleave 64-bit integers from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ pd Experimental Unpack and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠unpacklo_ ps Experimental Unpack and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠xor_ epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠xor_ epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst using zeromask k (elements are zeroed out when the corresponding mask bit is not set). - _mm_
maskz_ ⚠xor_ pd Experimental Compute the bitwise XOR of packed double-precision (64-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
maskz_ ⚠xor_ ps Experimental Compute the bitwise XOR of packed single-precision (32-bit) floating point numbers in a and b and store the results in dst using zeromask k (elements are zeroed out if the corresponding bit is not set). - _mm_
max_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed maximum values in dst. - _mm_
max_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed maximum values in dst. - _mm_
max_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed maximum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
max_ ⚠round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the maximum value in the lower element of dst, and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
max_ ⚠round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
max_ ⚠round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the maximum value in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
max_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the maximum value in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) maximum value when inputs are NaN or signed-zero values. - _mm_
min_ ⚠epi64 Experimental Compare packed signed 64-bit integers in a and b, and store packed minimum values in dst. - _mm_
min_ ⚠epu64 Experimental Compare packed unsigned 64-bit integers in a and b, and store packed minimum values in dst. - _mm_
min_ ⚠ph Experimental Compare packed half-precision (16-bit) floating-point elements in a and b, and store packed minimum values in dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
min_ ⚠round_ sd Experimental Compare the lower double-precision (64-bit) floating-point elements in a and b, store the minimum value in the lower element of dst , and copy the upper element from a to the upper element of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
min_ ⚠round_ sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
min_ ⚠round_ ss Experimental Compare the lower single-precision (32-bit) floating-point elements in a and b, store the minimum value in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
min_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b, store the minimum value in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Does not follow the IEEE Standard for Floating-Point Arithmetic (IEEE 754) minimum value when inputs are NaN or signed-zero values. - _mm_
mmask_ ⚠i32gather_ epi32 Experimental Loads 4 32-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i32gather_ epi64 Experimental Loads 2 64-bit integer elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i32gather_ pd Experimental Loads 2 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i32gather_ ps Experimental Loads 4 single-precision (32-bit) floating-point elements from memory starting at location base_addr at packed 32-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i64gather_ epi32 Experimental Loads 2 32-bit integer elements from memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i64gather_ epi64 Experimental Loads 2 64-bit integer elements from memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i64gather_ pd Experimental Loads 2 double-precision (64-bit) floating-point elements from memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
mmask_ ⚠i64gather_ ps Experimental Loads 2 single-precision (32-bit) floating-point elements from memory starting at location base_addr at packed 64-bit integer indices stored in vindex scaled by scale using writemask k (elements are copied from src when the corresponding mask bit is not set). - _mm_
move_ ⚠sh Experimental Move the lower half-precision (16-bit) floating-point element from b to the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
movepi8_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 8-bit integer in a. - _mm_
movepi16_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 16-bit integer in a. - _mm_
movepi32_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 32-bit integer in a. - _mm_
movepi64_ ⚠mask Experimental Set each bit of mask register k based on the most significant bit of the corresponding packed 64-bit integer in a. - _mm_
movm_ ⚠epi8 Experimental Set each packed 8-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm_
movm_ ⚠epi16 Experimental Set each packed 16-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm_
movm_ ⚠epi32 Experimental Set each packed 32-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm_
movm_ ⚠epi64 Experimental Set each packed 64-bit integer in dst to all ones or all zeros based on the value of the corresponding bit in k. - _mm_
mul_ ⚠pch Experimental Multiply packed complex numbers in a and b, and store the results in dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mul_ ⚠ph Experimental Multiply packed half-precision (16-bit) floating-point elements in a and b, and store the results in dst. - _mm_
mul_ ⚠round_ sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mul_ ⚠round_ sd Experimental Multiply the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
mul_ ⚠round_ sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
mul_ ⚠round_ ss Experimental Multiply the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
mul_ ⚠sch Experimental Multiply the lower complex numbers in a and b, and store the result in the lower elements of dst, and copy the upper 6 packed elements from a to the upper elements of dst. Each complex number is composed of two adjacent half-precision (16-bit) floating-point elements, which defines the complex numbercomplex = vec.fp16[0] + i * vec.fp16[1]
. - _mm_
mul_ ⚠sh Experimental Multiply the lower half-precision (16-bit) floating-point elements in a and b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
mullo_ ⚠epi64 Experimental Multiply packed 64-bit integers ina
andb
, producing intermediate 128-bit integers, and store the low 64 bits of the intermediate integers indst
. - _mm_
multishift_ ⚠epi64_ epi8 Experimental For each 64-bit element in b, select 8 unaligned bytes using a byte-granular shift control within the corresponding 64-bit element of a, and store the 8 assembled bytes to the corresponding 64-bit element of dst. - _mm_
or_ ⚠epi32 Experimental Compute the bitwise OR of packed 32-bit integers in a and b, and store the results in dst. - _mm_
or_ ⚠epi64 Experimental Compute the bitwise OR of packed 64-bit integers in a and b, and store the resut in dst. - _mm_
permutex2var_ ⚠epi8 Experimental Shuffle 8-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠epi16 Experimental Shuffle 16-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠epi32 Experimental Shuffle 32-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠epi64 Experimental Shuffle 64-bit integers in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠pd Experimental Shuffle double-precision (64-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a and b using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutex2var_ ⚠ps Experimental Shuffle single-precision (32-bit) floating-point elements in a and b across lanes using the corresponding selector and index in idx, and store the results in dst. - _mm_
permutexvar_ ⚠epi8 Experimental Shuffle 8-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm_
permutexvar_ ⚠epi16 Experimental Shuffle 16-bit integers in a across lanes using the corresponding index in idx, and store the results in dst. - _mm_
permutexvar_ ⚠ph Experimental Shuffle half-precision (16-bit) floating-point elements in a using the corresponding index in idx, and store the results in dst. - _mm_
popcnt_ ⚠epi8 Experimental For each packed 8-bit integer maps the value to the number of logical 1 bits. - _mm_
popcnt_ ⚠epi16 Experimental For each packed 16-bit integer maps the value to the number of logical 1 bits. - _mm_
popcnt_ ⚠epi32 Experimental For each packed 32-bit integer maps the value to the number of logical 1 bits. - _mm_
popcnt_ ⚠epi64 Experimental For each packed 64-bit integer maps the value to the number of logical 1 bits. - _mm_
range_ ⚠pd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed double-precision (64-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
range_ ⚠ps Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for packed single-precision (32-bit) floating-point elements in a and b, and store the results in dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. - _mm_
range_ ⚠round_ sd Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower double-precision (64-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
range_ ⚠round_ ss Experimental Calculate the max, min, absolute max, or absolute min (depending on control in imm8) for the lower single-precision (32-bit) floating-point element in a and b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. Lower 2 bits of IMM8 specifies the operation control: 00 = min, 01 = max, 10 = absolute min, 11 = absolute max. Upper 2 bits of IMM8 specifies the sign control: 00 = sign from a, 01 = sign from compare result, 10 = clear sign bit, 11 = set sign bit. Exceptions can be suppressed by passing _MM_FROUND_NO_EXC in the sae parameter. - _mm_
rcp14_ ⚠pd Experimental Compute the approximate reciprocal of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rcp14_ ⚠ps Experimental Compute the approximate reciprocal of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rcp14_ ⚠sd Experimental Compute the approximate reciprocal of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rcp14_ ⚠ss Experimental Compute the approximate reciprocal of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rcp_ ⚠ph Experimental Compute the approximate reciprocal of packed 16-bit floating-point elements ina
and stores the results indst
. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
rcp_ ⚠sh Experimental Compute the approximate reciprocal of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
reduce_ ⚠add_ epi8 Experimental Reduce the packed 8-bit integers in a by addition. Returns the sum of all elements in a. - _mm_
reduce_ ⚠add_ epi16 Experimental Reduce the packed 16-bit integers in a by addition. Returns the sum of all elements in a. - _mm_
reduce_ ⚠add_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by addition. Returns the sum of all elements in a. - _mm_
reduce_ ⚠and_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm_
reduce_ ⚠and_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise AND. Returns the bitwise AND of all elements in a. - _mm_
reduce_ ⚠max_ epi8 Experimental Reduce the packed 8-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm_
reduce_ ⚠max_ epi16 Experimental Reduce the packed 16-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm_
reduce_ ⚠max_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm_
reduce_ ⚠max_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by maximum. Returns the maximum of all elements in a. - _mm_
reduce_ ⚠max_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by maximum. Returns the maximum of all elements in a. - _mm_
reduce_ ⚠min_ epi8 Experimental Reduce the packed 8-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm_
reduce_ ⚠min_ epi16 Experimental Reduce the packed 16-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm_
reduce_ ⚠min_ epu8 Experimental Reduce the packed unsigned 8-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm_
reduce_ ⚠min_ epu16 Experimental Reduce the packed unsigned 16-bit integers in a by minimum. Returns the minimum of all elements in a. - _mm_
reduce_ ⚠min_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by minimum. Returns the minimum of all elements in a. - _mm_
reduce_ ⚠mul_ epi8 Experimental Reduce the packed 8-bit integers in a by multiplication. Returns the product of all elements in a. - _mm_
reduce_ ⚠mul_ epi16 Experimental Reduce the packed 16-bit integers in a by multiplication. Returns the product of all elements in a. - _mm_
reduce_ ⚠mul_ ph Experimental Reduce the packed half-precision (16-bit) floating-point elements in a by multiplication. Returns the product of all elements in a. - _mm_
reduce_ ⚠or_ epi8 Experimental Reduce the packed 8-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm_
reduce_ ⚠or_ epi16 Experimental Reduce the packed 16-bit integers in a by bitwise OR. Returns the bitwise OR of all elements in a. - _mm_
reduce_ ⚠pd Experimental Extract the reduced argument of packed double-precision (64-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
reduce_ ⚠ph Experimental Extract the reduced argument of packed half-precision (16-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. - _mm_
reduce_ ⚠ps Experimental Extract the reduced argument of packed single-precision (32-bit) floating-point elements in a by the number of bits specified by imm8, and store the results in dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
reduce_ ⚠round_ sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
reduce_ ⚠round_ sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
reduce_ ⚠round_ ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst, and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
reduce_ ⚠sd Experimental Extract the reduced argument of the lower double-precision (64-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst using, and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
reduce_ ⚠sh Experimental Extract the reduced argument of the lower half-precision (16-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
reduce_ ⚠ss Experimental Extract the reduced argument of the lower single-precision (32-bit) floating-point element in b by the number of bits specified by imm8, store the result in the lower element of dst, and copy the upper element from a. to the upper element of dst. Rounding is done according to the imm8 parameter, which can be one of: - _mm_
rol_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm_
rol_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in imm8, and store the results in dst. - _mm_
rolv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm_
rolv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the left by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm_
ror_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm_
ror_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in imm8, and store the results in dst. - _mm_
rorv_ ⚠epi32 Experimental Rotate the bits in each packed 32-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm_
rorv_ ⚠epi64 Experimental Rotate the bits in each packed 64-bit integer in a to the right by the number of bits specified in the corresponding element of b, and store the results in dst. - _mm_
roundscale_ ⚠pd Experimental Round packed double-precision (64-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
roundscale_ ⚠ph Experimental Round packed half-precision (16-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst. - _mm_
roundscale_ ⚠ps Experimental Round packed single-precision (32-bit) floating-point elements in a to the number of fraction bits specified by imm8, and store the results in dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
roundscale_ ⚠round_ sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
roundscale_ ⚠round_ sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
roundscale_ ⚠round_ ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE\ - _mm_
roundscale_ ⚠sd Experimental Round the lower double-precision (64-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
roundscale_ ⚠sh Experimental Round the lower half-precision (16-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
roundscale_ ⚠ss Experimental Round the lower single-precision (32-bit) floating-point element in b to the number of fraction bits specified by imm8, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.
Rounding is done according to the imm8[2:0] parameter, which can be one of:
_MM_FROUND_TO_NEAREST_INT // round to nearest
_MM_FROUND_TO_NEG_INF // round down
_MM_FROUND_TO_POS_INF // round up
_MM_FROUND_TO_ZERO // truncate
_MM_FROUND_CUR_DIRECTION // use MXCSR.RC; see _MM_SET_ROUNDING_MODE - _mm_
rsqrt14_ ⚠pd Experimental Compute the approximate reciprocal square root of packed double-precision (64-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rsqrt14_ ⚠ps Experimental Compute the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rsqrt14_ ⚠sd Experimental Compute the approximate reciprocal square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rsqrt14_ ⚠ss Experimental Compute the approximate reciprocal square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than 2^-14. - _mm_
rsqrt_ ⚠ph Experimental Compute the approximate reciprocal square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
rsqrt_ ⚠sh Experimental Compute the approximate reciprocal square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. The maximum relative error for this approximation is less than1.5*2^-12
. - _mm_
scalef_ ⚠pd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, and store the results in dst. - _mm_
scalef_ ⚠ph Experimental Scale the packed half-precision (16-bit) floating-point elements in a using values from b, and store the results in dst. - _mm_
scalef_ ⚠ps Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, and store the results in dst. - _mm_
scalef_ ⚠round_ sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
scalef_ ⚠round_ sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
scalef_ ⚠round_ ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
scalef_ ⚠sd Experimental Scale the packed double-precision (64-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst. - _mm_
scalef_ ⚠sh Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
scalef_ ⚠ss Experimental Scale the packed single-precision (32-bit) floating-point elements in a using values from b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst. - _mm_
set1_ ⚠ph Experimental Broadcast the half-precision (16-bit) floating-point value a to all elements of dst. - _mm_
set_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values. - _mm_
set_ ⚠sh Experimental Copy half-precision (16-bit) floating-point elements from a to the lower element of dst and zero the upper 7 elements. - _mm_
setr_ ⚠ph Experimental Set packed half-precision (16-bit) floating-point elements in dst with the supplied values in reverse order. - _mm_
setzero_ ⚠ph Experimental Return vector of type __m128h with all elements set to zero. - _mm_
shldi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by imm8 bits, and store the upper 16-bits in dst). - _mm_
shldi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by imm8 bits, and store the upper 32-bits in dst. - _mm_
shldi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by imm8 bits, and store the upper 64-bits in dst). - _mm_
shldv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in a and b producing an intermediate 32-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 16-bits in dst. - _mm_
shldv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in a and b producing an intermediate 64-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 32-bits in dst. - _mm_
shldv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in a and b producing an intermediate 128-bit result. Shift the result left by the amount specified in the corresponding element of c, and store the upper 64-bits in dst. - _mm_
shrdi_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by imm8 bits, and store the lower 16-bits in dst. - _mm_
shrdi_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by imm8 bits, and store the lower 32-bits in dst. - _mm_
shrdi_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by imm8 bits, and store the lower 64-bits in dst. - _mm_
shrdv_ ⚠epi16 Experimental Concatenate packed 16-bit integers in b and a producing an intermediate 32-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 16-bits in dst. - _mm_
shrdv_ ⚠epi32 Experimental Concatenate packed 32-bit integers in b and a producing an intermediate 64-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 32-bits in dst. - _mm_
shrdv_ ⚠epi64 Experimental Concatenate packed 64-bit integers in b and a producing an intermediate 128-bit result. Shift the result right by the amount specified in the corresponding element of c, and store the lower 64-bits in dst. - _mm_
sllv_ ⚠epi16 Experimental Shift packed 16-bit integers in a left by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm_
sqrt_ ⚠ph Experimental Compute the square root of packed half-precision (16-bit) floating-point elements in a, and store the results in dst. - _mm_
sqrt_ ⚠round_ sd Experimental Compute the square root of the lower double-precision (64-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
sqrt_ ⚠round_ sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
sqrt_ ⚠round_ ss Experimental Compute the square root of the lower single-precision (32-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
sqrt_ ⚠sh Experimental Compute the square root of the lower half-precision (16-bit) floating-point element in b, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
sra_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by count while shifting in sign bits, and store the results in dst. - _mm_
srai_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by imm8 while shifting in sign bits, and store the results in dst. - _mm_
srav_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm_
srav_ ⚠epi64 Experimental Shift packed 64-bit integers in a right by the amount specified by the corresponding element in count while shifting in sign bits, and store the results in dst. - _mm_
srlv_ ⚠epi16 Experimental Shift packed 16-bit integers in a right by the amount specified by the corresponding element in count while shifting in zeros, and store the results in dst. - _mm_
store_ ⚠epi32 Experimental Store 128-bits (composed of 4 packed 32-bit integers) from a into memory. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
store_ ⚠epi64 Experimental Store 128-bits (composed of 2 packed 64-bit integers) from a into memory. mem_addr must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_
store_ ⚠ph Experimental Store 128-bits (composed of 8 packed half-precision (16-bit) floating-point elements) from a into memory. The address must be aligned to 16 bytes or a general-protection exception may be generated. - _mm_
store_ ⚠sh Experimental Store the lower half-precision (16-bit) floating-point element from a into memory. - _mm_
storeu_ ⚠epi8 Experimental Store 128-bits (composed of 16 packed 8-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm_
storeu_ ⚠epi16 Experimental Store 128-bits (composed of 8 packed 16-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm_
storeu_ ⚠epi32 Experimental Store 128-bits (composed of 4 packed 32-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm_
storeu_ ⚠epi64 Experimental Store 128-bits (composed of 2 packed 64-bit integers) from a into memory. mem_addr does not need to be aligned on any particular boundary. - _mm_
storeu_ ⚠ph Experimental Store 128-bits (composed of 8 packed half-precision (16-bit) floating-point elements) from a into memory. The address does not need to be aligned to any particular boundary. - _mm_
sub_ ⚠ph Experimental Subtract packed half-precision (16-bit) floating-point elements in b from a, and store the results in dst. - _mm_
sub_ ⚠round_ sd Experimental Subtract the lower double-precision (64-bit) floating-point element in b from the lower double-precision (64-bit) floating-point element in a, store the result in the lower element of dst, and copy the upper element from a to the upper element of dst.\ - _mm_
sub_ ⚠round_ sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. Rounding is done according to the rounding parameter, which can be one of: - _mm_
sub_ ⚠round_ ss Experimental Subtract the lower single-precision (32-bit) floating-point element in b from the lower single-precision (32-bit) floating-point element in a, store the result in the lower element of dst, and copy the upper 3 packed elements from a to the upper elements of dst.\ - _mm_
sub_ ⚠sh Experimental Subtract the lower half-precision (16-bit) floating-point elements in b from a, store the result in the lower element of dst, and copy the upper 7 packed elements from a to the upper elements of dst. - _mm_
ternarylogic_ ⚠epi32 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 32-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm_
ternarylogic_ ⚠epi64 Experimental Bitwise ternary logic that provides the capability to implement any three-operand binary function; the specific binary function is specified by value in imm8. For each bit in each packed 64-bit integer, the corresponding bit from a, b, and c are used to form a 3 bit index into imm8, and the value at that bit in imm8 is written to the corresponding bit in dst. - _mm_
test_ ⚠epi8_ mask Experimental Compute the bitwise AND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm_
test_ ⚠epi16_ mask Experimental Compute the bitwise AND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm_
test_ ⚠epi32_ mask Experimental Compute the bitwise AND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm_
test_ ⚠epi64_ mask Experimental Compute the bitwise AND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is non-zero. - _mm_
testn_ ⚠epi8_ mask Experimental Compute the bitwise NAND of packed 8-bit integers in a and b, producing intermediate 8-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm_
testn_ ⚠epi16_ mask Experimental Compute the bitwise NAND of packed 16-bit integers in a and b, producing intermediate 16-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm_
testn_ ⚠epi32_ mask Experimental Compute the bitwise NAND of packed 32-bit integers in a and b, producing intermediate 32-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm_
testn_ ⚠epi64_ mask Experimental Compute the bitwise NAND of packed 64-bit integers in a and b, producing intermediate 64-bit values, and set the corresponding bit in result mask k if the intermediate value is zero. - _mm_
ucomieq_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for equality, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
ucomige_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for greater-than-or-equal, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
ucomigt_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for greater-than, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
ucomile_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for less-than-or-equal, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
ucomilt_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for less-than, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
ucomineq_ ⚠sh Experimental Compare the lower half-precision (16-bit) floating-point elements in a and b for not-equal, and return the boolean result (0 or 1). This instruction will not signal an exception for QNaNs. - _mm_
undefined_ ⚠ph Experimental Return vector of type__m128h
with undefined elements. In practice, this returns the all-zero vector. - _mm_
xor_ ⚠epi32 Experimental Compute the bitwise XOR of packed 32-bit integers in a and b, and store the results in dst. - _mm_
xor_ ⚠epi64 Experimental Compute the bitwise XOR of packed 64-bit integers in a and b, and store the results in dst. - _store_
mask8 ⚠Experimental Store 8-bit mask to memory - _store_
mask16 ⚠Experimental Store 16-bit mask to memory - _store_
mask32 ⚠Experimental Store 32-bit mask from a into memory. - _store_
mask64 ⚠Experimental Store 64-bit mask from a into memory. - _tile_
cmmimfp16ps ⚠Experimental Perform matrix multiplication of two tiles containing complex elements and accumulate the results into a packed single precision tile. Each dword element in input tiles a and b is interpreted as a complex number with FP16 real part and FP16 imaginary part. Calculates the imaginary part of the result. For each possible combination of (row of a, column of b), it performs a set of multiplication and accumulations on all corresponding complex numbers (one from a and one from b). The imaginary part of the a element is multiplied with the real part of the corresponding b element, and the real part of the a element is multiplied with the imaginary part of the corresponding b elements. The two accumulated results are added, and then accumulated into the corresponding row and column of dst. - _tile_
cmmrlfp16ps ⚠Experimental Perform matrix multiplication of two tiles containing complex elements and accumulate the results into a packed single precision tile. Each dword element in input tiles a and b is interpreted as a complex number with FP16 real part and FP16 imaginary part. Calculates the real part of the result. For each possible combination of (row of a, column of b), it performs a set of multiplication and accumulations on all corresponding complex numbers (one from a and one from b). The real part of the a element is multiplied with the real part of the corresponding b element, and the negated imaginary part of the a element is multiplied with the imaginary part of the corresponding b elements. The two accumulated results are added, and then accumulated into the corresponding row and column of dst. - _tile_
dpbf16ps ⚠Experimental Compute dot-product of BF16 (16-bit) floating-point pairs in tiles a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in dst, and store the 32-bit result back to tile dst. - _tile_
dpbssd ⚠Experimental Compute dot-product of bytes in tiles with a source/destination accumulator. Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate 32-bit results. Sum these 4 results with the corresponding 32-bit integer in dst, and store the 32-bit result back to tile dst. - _tile_
dpbsud ⚠Experimental Compute dot-product of bytes in tiles with a source/destination accumulator. Multiply groups of 4 adjacent pairs of signed 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate 32-bit results. Sum these 4 results with the corresponding 32-bit integer in dst, and store the 32-bit result back to tile dst. - _tile_
dpbusd ⚠Experimental Compute dot-product of bytes in tiles with a source/destination accumulator. Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding signed 8-bit integers in b, producing 4 intermediate 32-bit results. Sum these 4 results with the corresponding 32-bit integer in dst, and store the 32-bit result back to tile dst. - _tile_
dpbuud ⚠Experimental Compute dot-product of bytes in tiles with a source/destination accumulator. Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in a with corresponding unsigned 8-bit integers in b, producing 4 intermediate 32-bit results. Sum these 4 results with the corresponding 32-bit integer in dst, and store the 32-bit result back to tile dst. - _tile_
dpfp16ps ⚠Experimental Compute dot-product of FP16 (16-bit) floating-point pairs in tiles a and b, accumulating the intermediate single-precision (32-bit) floating-point elements with elements in dst, and store the 32-bit result back to tile dst. - _tile_
loadconfig ⚠Experimental Load tile configuration from a 64-byte memory location specified by mem_addr. The tile configuration format is specified below, and includes the tile type pallette, the number of bytes per row, and the number of rows. If the specified pallette_id is zero, that signifies the init state for both the tile config and the tile data, and the tiles are zeroed. Any invalid configurations will result in #GP fault. - _tile_
loadd ⚠Experimental Load tile rows from memory specifieid by base address and stride into destination tile dst using the tile configuration previously configured via _tile_loadconfig. - _tile_
release ⚠Experimental Release the tile configuration to return to the init state, which releases all storage it currently holds. - _tile_
storeconfig ⚠Experimental Stores the current tile configuration to a 64-byte memory location specified by mem_addr. The tile configuration format is specified below, and includes the tile type pallette, the number of bytes per row, and the number of rows. If tiles are not configured, all zeroes will be stored to memory. - _tile_
stored ⚠Experimental Store the tile specified by src to memory specifieid by base address and stride using the tile configuration previously configured via _tile_loadconfig. - _tile_
stream_ ⚠loadd Experimental Load tile rows from memory specifieid by base address and stride into destination tile dst using the tile configuration previously configured via _tile_loadconfig. This intrinsic provides a hint to the implementation that the data will likely not be reused in the near future and the data caching can be optimized accordingly. - _tile_
zero ⚠Experimental Zero the tile specified by tdest. - _xabort⚠
Experimental Forces a restricted transactional memory (RTM) region to abort. - _xabort_
code Experimental - _xbegin⚠
Experimental Specifies the start of a restricted transactional memory (RTM) code region and returns a value indicating status. - _xend⚠
Experimental Specifies the end of a restricted transactional memory (RTM) code region. - _xtest⚠
Experimental Queries whether the processor is executing in a transactional region identified by restricted transactional memory (RTM) or hardware lock elision (HLE).
Type Aliases§
- _MM_
CMPINT_ ENUM Experimental The_MM_CMPINT_ENUM
type used to specify comparison operations in AVX-512 intrinsics. - _MM_
MANTISSA_ NORM_ ENUM Experimental TheMM_MANTISSA_NORM_ENUM
type used to specify mantissa normalized operations in AVX-512 intrinsics. - _MM_
MANTISSA_ SIGN_ ENUM Experimental TheMM_MANTISSA_SIGN_ENUM
type used to specify mantissa signed operations in AVX-512 intrinsics. - _MM_
PERM_ ENUM Experimental TheMM_PERM_ENUM
type used to specify shuffle operations in AVX-512 intrinsics. - __
mmask8 Experimental The__mmask8
type used in AVX-512 intrinsics, a 8-bit integer - __
mmask16 Experimental The__mmask16
type used in AVX-512 intrinsics, a 16-bit integer - __
mmask32 Experimental The__mmask32
type used in AVX-512 intrinsics, a 32-bit integer - __
mmask64 Experimental The__mmask64
type used in AVX-512 intrinsics, a 64-bit integer