Available on x86 only.
Expand description
Platform-specific intrinsics for the x86
platform.
See the module documentation for more details.
Structs§
- CpuidResultx86 or x86-64Result of the
cpuid
instruction. - __m128x86 or x86-64128-bit wide set of four
f32
types, x86-specific - __m256x86 or x86-64256-bit wide set of eight
f32
types, x86-specific - __m512x86 or x86-64512-bit wide set of sixteen
f32
types, x86-specific - __m128dx86 or x86-64128-bit wide set of two
f64
types, x86-specific - __m128ix86 or x86-64128-bit wide integer vector type, x86-specific
- __m256dx86 or x86-64256-bit wide set of four
f64
types, x86-specific - __m256ix86 or x86-64256-bit wide integer vector type, x86-specific
- __m512dx86 or x86-64512-bit wide set of eight
f64
types, x86-specific - __m512ix86 or x86-64512-bit wide integer vector type, x86-specific
- 128-bit wide set of eight
u16
types, x86-specific - 256-bit wide set of 16
u16
types, x86-specific - 512-bit wide set of 32
u16
types, x86-specific
Constants§
- _CMP_EQ_OQx86 or x86-64Equal (ordered, non-signaling)
- _CMP_EQ_OSx86 or x86-64Equal (ordered, signaling)
- _CMP_EQ_UQx86 or x86-64Equal (unordered, non-signaling)
- _CMP_EQ_USx86 or x86-64Equal (unordered, signaling)
- _CMP_FALSE_OQx86 or x86-64False (ordered, non-signaling)
- _CMP_FALSE_OSx86 or x86-64False (ordered, signaling)
- _CMP_GE_OQx86 or x86-64Greater-than-or-equal (ordered, non-signaling)
- _CMP_GE_OSx86 or x86-64Greater-than-or-equal (ordered, signaling)
- _CMP_GT_OQx86 or x86-64Greater-than (ordered, non-signaling)
- _CMP_GT_OSx86 or x86-64Greater-than (ordered, signaling)
- _CMP_LE_OQx86 or x86-64Less-than-or-equal (ordered, non-signaling)
- _CMP_LE_OSx86 or x86-64Less-than-or-equal (ordered, signaling)
- _CMP_LT_OQx86 or x86-64Less-than (ordered, non-signaling)
- _CMP_LT_OSx86 or x86-64Less-than (ordered, signaling)
- _CMP_NEQ_OQx86 or x86-64Not-equal (ordered, non-signaling)
- _CMP_NEQ_OSx86 or x86-64Not-equal (ordered, signaling)
- _CMP_NEQ_UQx86 or x86-64Not-equal (unordered, non-signaling)
- _CMP_NEQ_USx86 or x86-64Not-equal (unordered, signaling)
- _CMP_NGE_UQx86 or x86-64Not-greater-than-or-equal (unordered, non-signaling)
- _CMP_NGE_USx86 or x86-64Not-greater-than-or-equal (unordered, signaling)
- _CMP_NGT_UQx86 or x86-64Not-greater-than (unordered, non-signaling)
- _CMP_NGT_USx86 or x86-64Not-greater-than (unordered, signaling)
- _CMP_NLE_UQx86 or x86-64Not-less-than-or-equal (unordered, non-signaling)
- _CMP_NLE_USx86 or x86-64Not-less-than-or-equal (unordered, signaling)
- _CMP_NLT_UQx86 or x86-64Not-less-than (unordered, non-signaling)
- _CMP_NLT_USx86 or x86-64Not-less-than (unordered, signaling)
- _CMP_ORD_Qx86 or x86-64Ordered (non-signaling)
- _CMP_ORD_Sx86 or x86-64Ordered (signaling)
- _CMP_TRUE_UQx86 or x86-64True (unordered, non-signaling)
- _CMP_TRUE_USx86 or x86-64True (unordered, signaling)
- _CMP_UNORD_Qx86 or x86-64Unordered (non-signaling)
- _CMP_UNORD_Sx86 or x86-64Unordered (signaling)
- _MM_EXCEPT_DENORMx86 or x86-64See
_mm_setcsr
- _MM_EXCEPT_DIV_ZEROx86 or x86-64See
_mm_setcsr
- _MM_EXCEPT_INEXACTx86 or x86-64See
_mm_setcsr
- _MM_EXCEPT_INVALIDx86 or x86-64See
_mm_setcsr
- _MM_EXCEPT_MASKx86 or x86-64
- _MM_EXCEPT_OVERFLOWx86 or x86-64See
_mm_setcsr
- _MM_EXCEPT_UNDERFLOWx86 or x86-64See
_mm_setcsr
- _MM_FLUSH_ZERO_MASKx86 or x86-64
- _MM_FLUSH_ZERO_OFFx86 or x86-64See
_mm_setcsr
- _MM_FLUSH_ZERO_ONx86 or x86-64See
_mm_setcsr
- _MM_FROUND_CEILx86 or x86-64round up and do not suppress exceptions
- _MM_FROUND_CUR_DIRECTIONx86 or x86-64use MXCSR.RC; see
vendor::_MM_SET_ROUNDING_MODE
- _MM_FROUND_FLOORx86 or x86-64round down and do not suppress exceptions
- _MM_FROUND_NEARBYINTx86 or x86-64use MXCSR.RC and suppress exceptions; see
vendor::_MM_SET_ROUNDING_MODE
- _MM_FROUND_NINTx86 or x86-64round to nearest and do not suppress exceptions
- _MM_FROUND_NO_EXCx86 or x86-64suppress exceptions
- _MM_FROUND_RAISE_EXCx86 or x86-64do not suppress exceptions
- _MM_FROUND_RINTx86 or x86-64use MXCSR.RC and do not suppress exceptions; see
vendor::_MM_SET_ROUNDING_MODE
- _MM_FROUND_TO_NEAREST_INTx86 or x86-64round to nearest
- _MM_FROUND_TO_NEG_INFx86 or x86-64round down
- _MM_FROUND_TO_POS_INFx86 or x86-64round up
- _MM_FROUND_TO_ZEROx86 or x86-64truncate
- _MM_FROUND_TRUNCx86 or x86-64truncate and do not suppress exceptions
- _MM_HINT_ET0x86 or x86-64See
_mm_prefetch
. - _MM_HINT_ET1x86 or x86-64See
_mm_prefetch
. - _MM_HINT_NTAx86 or x86-64See
_mm_prefetch
. - _MM_HINT_T0x86 or x86-64See
_mm_prefetch
. - _MM_HINT_T1x86 or x86-64See
_mm_prefetch
. - _MM_HINT_T2x86 or x86-64See
_mm_prefetch
. - _MM_MASK_DENORMx86 or x86-64See
_mm_setcsr
- _MM_MASK_DIV_ZEROx86 or x86-64See
_mm_setcsr
- _MM_MASK_INEXACTx86 or x86-64See
_mm_setcsr
- _MM_MASK_INVALIDx86 or x86-64See
_mm_setcsr
- _MM_MASK_MASKx86 or x86-64
- _MM_MASK_OVERFLOWx86 or x86-64See
_mm_setcsr
- _MM_MASK_UNDERFLOWx86 or x86-64See
_mm_setcsr
- _MM_ROUND_DOWNx86 or x86-64See
_mm_setcsr
- _MM_ROUND_MASKx86 or x86-64
- _MM_ROUND_NEARESTx86 or x86-64See
_mm_setcsr
- _MM_ROUND_TOWARD_ZEROx86 or x86-64See
_mm_setcsr
- _MM_ROUND_UPx86 or x86-64See
_mm_setcsr
- _SIDD_BIT_MASKx86 or x86-64Mask only: return the bit mask
- _SIDD_CMP_EQUAL_ANYx86 or x86-64For each character in
a
, find if it is inb
(Default) - _SIDD_CMP_EQUAL_EACHx86 or x86-64The strings defined by
a
andb
are equal - _SIDD_CMP_EQUAL_ORDEREDx86 or x86-64Search for the defined substring in the target
- _SIDD_CMP_RANGESx86 or x86-64For each character in
a
, determine ifb[0] <= c <= b[1] or b[1] <= c <= b[2]...
- _SIDD_LEAST_SIGNIFICANTx86 or x86-64Index only: return the least significant bit (Default)
- _SIDD_MASKED_NEGATIVE_POLARITYx86 or x86-64Negates results only before the end of the string
- _SIDD_MASKED_POSITIVE_POLARITYx86 or x86-64Do not negate results before the end of the string
- _SIDD_MOST_SIGNIFICANTx86 or x86-64Index only: return the most significant bit
- _SIDD_NEGATIVE_POLARITYx86 or x86-64Negates results
- _SIDD_POSITIVE_POLARITYx86 or x86-64Do not negate results (Default)
- _SIDD_SBYTE_OPSx86 or x86-64String contains signed 8-bit characters
- _SIDD_SWORD_OPSx86 or x86-64String contains unsigned 16-bit characters
- _SIDD_UBYTE_OPSx86 or x86-64String contains unsigned 8-bit characters (Default)
- _SIDD_UNIT_MASKx86 or x86-64Mask only: return the byte mask
- _SIDD_UWORD_OPSx86 or x86-64String contains unsigned 16-bit characters
- _XCR_XFEATURE_ENABLED_MASKx86 or x86-64
XFEATURE_ENABLED_MASK
forXCR
- Equal
- False
- Less-than-or-equal
- Less-than
- Not-equal
- Not less-than-or-equal
- Not less-than
- True
- interval [1, 2)
- interval [0.5, 1)
- interval [0.5, 2)
- interval [0.75, 1.5)
- DEST = NaN if sign(SRC) = 1
- sign = sign(SRC)
- sign = 0
- Transaction abort due to the transaction using too much memory.
- Transaction abort due to a memory conflict with another thread.
- Transaction abort due to a debug trap.
- Transaction explicitly aborted with xabort. The parameter passed to xabort is available with
_xabort_code(status)
. - Transaction abort in a inner nested transaction.
- Transaction retry is possible.
- Transaction successfully started.
Functions§
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- See
_mm_setcsr
- _MM_TRANSPOSE4_PS⚠(x86 or x86-64) and
sse
Transpose the 4x4 matrix formed by 4 rows of __m128 in place. - __cpuid⚠x86 or x86-64See
__cpuid_count
. - __cpuid_count⚠x86 or x86-64Returns the result of the
cpuid
instruction for a givenleaf
(EAX
) andsub_leaf
(ECX
). - __get_cpuid_max⚠x86 or x86-64Returns the highest-supported
leaf
(EAX
) and sub-leaf (ECX
)cpuid
values. - __rdtscp⚠x86 or x86-64Reads the current value of the processor’s time-stamp counter and the
IA32_TSC_AUX MSR
. - _addcarry_u32⚠x86 or x86-64Adds unsigned 32-bit integers
a
andb
with unsigned 8-bit carry-inc_in
(carry flag), and store the unsigned 32-bit result inout
, and the carry-out is returned (carry or overflow flag). - _addcarryx_u32⚠(x86 or x86-64) and
adx
Adds unsigned 32-bit integersa
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). - _andn_u32⚠(x86 or x86-64) and
bmi1
Bitwise logicalAND
of inverteda
withb
. - _bextr2_u32⚠(x86 or x86-64) and
bmi1
Extracts bits ofa
specified bycontrol
into the least significant bits of the result. - _bextr_u32⚠(x86 or x86-64) and
bmi1
Extracts bits in range [start
,start
+length
) froma
into the least significant bits of the result. - _bittest⚠x86 or x86-64Returns the bit in position
b
of the memory addressed byp
. - _bittestandcomplement⚠x86 or x86-64Returns the bit in position
b
of the memory addressed byp
, then inverts that bit. - _bittestandreset⚠x86 or x86-64Returns the bit in position
b
of the memory addressed byp
, then resets that bit to0
. - _bittestandset⚠x86 or x86-64Returns the bit in position
b
of the memory addressed byp
, then sets the bit to1
. - _blcfill_u32⚠(x86 or x86-64) and
tbm
Clears all bits below the least significant zero bit ofx
. - _blcfill_u64⚠(x86 or x86-64) and
tbm
Clears all bits below the least significant zero bit ofx
. - _blci_u32⚠(x86 or x86-64) and
tbm
Sets all bits ofx
to 1 except for the least significant zero bit. - _blci_u64⚠(x86 or x86-64) and
tbm
Sets all bits ofx
to 1 except for the least significant zero bit. - _blcic_u32⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
and clears all other bits. - _blcic_u64⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
and clears all other bits. - _blcmsk_u32⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
and clears all bits above that bit. - _blcmsk_u64⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
and clears all bits above that bit. - _blcs_u32⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
. - _blcs_u64⚠(x86 or x86-64) and
tbm
Sets the least significant zero bit ofx
. - _blsfill_u32⚠(x86 or x86-64) and
tbm
Sets all bits ofx
below the least significant one. - _blsfill_u64⚠(x86 or x86-64) and
tbm
Sets all bits ofx
below the least significant one. - _blsi_u32⚠(x86 or x86-64) and
bmi1
Extracts lowest set isolated bit. - _blsic_u32⚠(x86 or x86-64) and
tbm
Clears least significant bit and sets all other bits. - _blsic_u64⚠(x86 or x86-64) and
tbm
Clears least significant bit and sets all other bits. - _blsmsk_u32⚠(x86 or x86-64) and
bmi1
Gets mask up to lowest set bit. - _blsr_u32⚠(x86 or x86-64) and
bmi1
Resets the lowest set bit ofx
. - _bswap⚠x86 or x86-64Returns an integer with the reversed byte order of x
- _bzhi_u32⚠(x86 or x86-64) and
bmi2
Zeroes higher bits ofa
>=index
. - _fxrstor⚠(x86 or x86-64) and
fxsr
Restores theXMM
,MMX
,MXCSR
, andx87
FPU registers from the 512-byte-long 16-byte-aligned memory regionmem_addr
. - _fxsave⚠(x86 or x86-64) and
fxsr
Saves thex87
FPU,MMX
technology,XMM
, andMXCSR
registers to the 512-byte-long 16-byte-aligned memory regionmem_addr
. - _lzcnt_u32⚠(x86 or x86-64) and
lzcnt
Counts the leading most significant zero bits. - _mm256_abs_epi8⚠(x86 or x86-64) and
avx2
Computes the absolute values of packed 8-bit integers ina
. - _mm256_abs_epi16⚠(x86 or x86-64) and
avx2
Computes the absolute values of packed 16-bit integers ina
. - _mm256_abs_epi32⚠(x86 or x86-64) and
avx2
Computes the absolute values of packed 32-bit integers ina
. - _mm256_add_epi8⚠(x86 or x86-64) and
avx2
Adds packed 8-bit integers ina
andb
. - _mm256_add_epi16⚠(x86 or x86-64) and
avx2
Adds packed 16-bit integers ina
andb
. - _mm256_add_epi32⚠(x86 or x86-64) and
avx2
Adds packed 32-bit integers ina
andb
. - _mm256_add_epi64⚠(x86 or x86-64) and
avx2
Adds packed 64-bit integers ina
andb
. - _mm256_add_pd⚠(x86 or x86-64) and
avx
Adds packed double-precision (64-bit) floating-point elements ina
andb
. - _mm256_add_ps⚠(x86 or x86-64) and
avx
Adds packed single-precision (32-bit) floating-point elements ina
andb
. - _mm256_adds_epi8⚠(x86 or x86-64) and
avx2
Adds packed 8-bit integers ina
andb
using saturation. - _mm256_adds_epi16⚠(x86 or x86-64) and
avx2
Adds packed 16-bit integers ina
andb
using saturation. - _mm256_adds_epu8⚠(x86 or x86-64) and
avx2
Adds packed unsigned 8-bit integers ina
andb
using saturation. - _mm256_adds_epu16⚠(x86 or x86-64) and
avx2
Adds packed unsigned 16-bit integers ina
andb
using saturation. - _mm256_addsub_pd⚠(x86 or x86-64) and
avx
Alternatively adds and subtracts packed double-precision (64-bit) floating-point elements ina
to/from packed elements inb
. - _mm256_addsub_ps⚠(x86 or x86-64) and
avx
Alternatively adds and subtracts packed single-precision (32-bit) floating-point elements ina
to/from packed elements inb
. - _mm256_alignr_epi8⚠(x86 or x86-64) and
avx2
Concatenates pairs of 16-byte blocks ina
andb
into a 32-byte temporary result, shifts the result right byn
bytes, and returns the low 16 bytes. - _mm256_and_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of a packed double-precision (64-bit) floating-point elements ina
andb
. - _mm256_and_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of packed single-precision (32-bit) floating-point elements ina
andb
. - _mm256_and_si256⚠(x86 or x86-64) and
avx2
Computes the bitwise AND of 256 bits (representing integer data) ina
andb
. - _mm256_andnot_pd⚠(x86 or x86-64) and
avx
Computes the bitwise NOT of packed double-precision (64-bit) floating-point elements ina
, and then AND withb
. - _mm256_andnot_ps⚠(x86 or x86-64) and
avx
Computes the bitwise NOT of packed single-precision (32-bit) floating-point elements ina
and then AND withb
. - _mm256_andnot_si256⚠(x86 or x86-64) and
avx2
Computes the bitwise NOT of 256 bits (representing integer data) ina
and then AND withb
. - _mm256_avg_epu8⚠(x86 or x86-64) and
avx2
Averages packed unsigned 8-bit integers ina
andb
. - _mm256_avg_epu16⚠(x86 or x86-64) and
avx2
Averages packed unsigned 16-bit integers ina
andb
. - _mm256_blend_epi16⚠(x86 or x86-64) and
avx2
Blends packed 16-bit integers froma
andb
using control maskIMM8
. - _mm256_blend_epi32⚠(x86 or x86-64) and
avx2
Blends packed 32-bit integers froma
andb
using control maskIMM8
. - _mm256_blend_pd⚠(x86 or x86-64) and
avx
Blends packed double-precision (64-bit) floating-point elements froma
andb
using control maskimm8
. - _mm256_blend_ps⚠(x86 or x86-64) and
avx
Blends packed single-precision (32-bit) floating-point elements froma
andb
using control maskimm8
. - _mm256_blendv_epi8⚠(x86 or x86-64) and
avx2
Blends packed 8-bit integers froma
andb
usingmask
. - _mm256_blendv_pd⚠(x86 or x86-64) and
avx
Blends packed double-precision (64-bit) floating-point elements froma
andb
usingc
as a mask. - _mm256_blendv_ps⚠(x86 or x86-64) and
avx
Blends packed single-precision (32-bit) floating-point elements froma
andb
usingc
as a mask. - _mm256_broadcast_pd⚠(x86 or x86-64) and
avx
Broadcasts 128 bits from memory (composed of 2 packed double-precision (64-bit) floating-point elements) to all elements of the returned vector. - _mm256_broadcast_ps⚠(x86 or x86-64) and
avx
Broadcasts 128 bits from memory (composed of 4 packed single-precision (32-bit) floating-point elements) to all elements of the returned vector. - _mm256_broadcast_sd⚠(x86 or x86-64) and
avx
Broadcasts a double-precision (64-bit) floating-point element from memory to all elements of the returned vector. - _mm256_broadcast_ss⚠(x86 or x86-64) and
avx
Broadcasts a single-precision (32-bit) floating-point element from memory to all elements of the returned vector. - _mm256_broadcastb_epi8⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 8-bit integer froma
to all elements of the 256-bit returned value. - _mm256_broadcastd_epi32⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 32-bit integer froma
to all elements of the 256-bit returned value. - _mm256_broadcastq_epi64⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 64-bit integer froma
to all elements of the 256-bit returned value. - _mm256_broadcastsd_pd⚠(x86 or x86-64) and
avx2
Broadcasts the low double-precision (64-bit) floating-point element froma
to all elements of the 256-bit returned value. - _mm256_broadcastsi128_si256⚠(x86 or x86-64) and
avx2
Broadcasts 128 bits of integer data from a to all 128-bit lanes in the 256-bit returned value. - _mm256_broadcastss_ps⚠(x86 or x86-64) and
avx2
Broadcasts the low single-precision (32-bit) floating-point element froma
to all elements of the 256-bit returned value. - _mm256_broadcastw_epi16⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 16-bit integer from a to all elements of the 256-bit returned value - _mm256_bslli_epi128⚠(x86 or x86-64) and
avx2
Shifts 128-bit lanes ina
left byimm8
bytes while shifting in zeros. - _mm256_bsrli_epi128⚠(x86 or x86-64) and
avx2
Shifts 128-bit lanes ina
right byimm8
bytes while shifting in zeros. - _mm256_castpd128_pd256⚠(x86 or x86-64) and
avx
Casts vector of type __m128d to type __m256d; the upper 128 bits of the result are undefined. - _mm256_castpd256_pd128⚠(x86 or x86-64) and
avx
Casts vector of type __m256d to type __m128d. - _mm256_castpd_ps⚠(x86 or x86-64) and
avx
Cast vector of type __m256d to type __m256. - _mm256_castpd_si256⚠(x86 or x86-64) and
avx
Casts vector of type __m256d to type __m256i. - _mm256_castps128_ps256⚠(x86 or x86-64) and
avx
Casts vector of type __m128 to type __m256; the upper 128 bits of the result are undefined. - _mm256_castps256_ps128⚠(x86 or x86-64) and
avx
Casts vector of type __m256 to type __m128. - _mm256_castps_pd⚠(x86 or x86-64) and
avx
Cast vector of type __m256 to type __m256d. - _mm256_castps_si256⚠(x86 or x86-64) and
avx
Casts vector of type __m256 to type __m256i. - _mm256_castsi128_si256⚠(x86 or x86-64) and
avx
Casts vector of type __m128i to type __m256i; the upper 128 bits of the result are undefined. - _mm256_castsi256_pd⚠(x86 or x86-64) and
avx
Casts vector of type __m256i to type __m256d. - _mm256_castsi256_ps⚠(x86 or x86-64) and
avx
Casts vector of type __m256i to type __m256. - _mm256_castsi256_si128⚠(x86 or x86-64) and
avx
Casts vector of type __m256i to type __m128i. - _mm256_ceil_pd⚠(x86 or x86-64) and
avx
Rounds packed double-precision (64-bit) floating point elements ina
toward positive infinity. - _mm256_ceil_ps⚠(x86 or x86-64) and
avx
Rounds packed single-precision (32-bit) floating point elements ina
toward positive infinity. - _mm256_cmp_pd⚠(x86 or x86-64) and
avx
Compares packed double-precision (64-bit) floating-point elements ina
andb
based on the comparison operand specified byIMM5
. - _mm256_cmp_ps⚠(x86 or x86-64) and
avx
Compares packed single-precision (32-bit) floating-point elements ina
andb
based on the comparison operand specified byIMM5
. - _mm256_cmpeq_epi8⚠(x86 or x86-64) and
avx2
Compares packed 8-bit integers ina
andb
for equality. - _mm256_cmpeq_epi16⚠(x86 or x86-64) and
avx2
Compares packed 16-bit integers ina
andb
for equality. - _mm256_cmpeq_epi32⚠(x86 or x86-64) and
avx2
Compares packed 32-bit integers ina
andb
for equality. - _mm256_cmpeq_epi64⚠(x86 or x86-64) and
avx2
Compares packed 64-bit integers ina
andb
for equality. - _mm256_cmpgt_epi8⚠(x86 or x86-64) and
avx2
Compares packed 8-bit integers ina
andb
for greater-than. - _mm256_cmpgt_epi16⚠(x86 or x86-64) and
avx2
Compares packed 16-bit integers ina
andb
for greater-than. - _mm256_cmpgt_epi32⚠(x86 or x86-64) and
avx2
Compares packed 32-bit integers ina
andb
for greater-than. - _mm256_cmpgt_epi64⚠(x86 or x86-64) and
avx2
Compares packed 64-bit integers ina
andb
for greater-than. - _mm256_cvtepi8_epi16⚠(x86 or x86-64) and
avx2
Sign-extend 8-bit integers to 16-bit integers. - _mm256_cvtepi8_epi32⚠(x86 or x86-64) and
avx2
Sign-extend 8-bit integers to 32-bit integers. - _mm256_cvtepi8_epi64⚠(x86 or x86-64) and
avx2
Sign-extend 8-bit integers to 64-bit integers. - _mm256_cvtepi16_epi32⚠(x86 or x86-64) and
avx2
Sign-extend 16-bit integers to 32-bit integers. - _mm256_cvtepi16_epi64⚠(x86 or x86-64) and
avx2
Sign-extend 16-bit integers to 64-bit integers. - _mm256_cvtepi32_epi64⚠(x86 or x86-64) and
avx2
Sign-extend 32-bit integers to 64-bit integers. - _mm256_cvtepi32_pd⚠(x86 or x86-64) and
avx
Converts packed 32-bit integers ina
to packed double-precision (64-bit) floating-point elements. - _mm256_cvtepi32_ps⚠(x86 or x86-64) and
avx
Converts packed 32-bit integers ina
to packed single-precision (32-bit) floating-point elements. - _mm256_cvtepu8_epi16⚠(x86 or x86-64) and
avx2
Zero-extend unsigned 8-bit integers ina
to 16-bit integers. - _mm256_cvtepu8_epi32⚠(x86 or x86-64) and
avx2
Zero-extend the lower eight unsigned 8-bit integers ina
to 32-bit integers. The upper eight elements ofa
are unused. - _mm256_cvtepu8_epi64⚠(x86 or x86-64) and
avx2
Zero-extend the lower four unsigned 8-bit integers ina
to 64-bit integers. The upper twelve elements ofa
are unused. - _mm256_cvtepu16_epi32⚠(x86 or x86-64) and
avx2
Zeroes extend packed unsigned 16-bit integers ina
to packed 32-bit integers, and stores the results indst
. - _mm256_cvtepu16_epi64⚠(x86 or x86-64) and
avx2
Zero-extend the lower four unsigned 16-bit integers ina
to 64-bit integers. The upper four elements ofa
are unused. - _mm256_cvtepu32_epi64⚠(x86 or x86-64) and
avx2
Zero-extend unsigned 32-bit integers ina
to 64-bit integers. - _mm256_cvtpd_epi32⚠(x86 or x86-64) and
avx
Converts packed double-precision (64-bit) floating-point elements ina
to packed 32-bit integers. - _mm256_cvtpd_ps⚠(x86 or x86-64) and
avx
Converts packed double-precision (64-bit) floating-point elements ina
to packed single-precision (32-bit) floating-point elements. - _mm256_cvtph_ps⚠(x86 or x86-64) and
f16c
Converts the 8 x 16-bit half-precision float values in the 128-bit vectora
into 8 x 32-bit float values stored in a 256-bit wide vector. - _mm256_cvtps_epi32⚠(x86 or x86-64) and
avx
Converts packed single-precision (32-bit) floating-point elements ina
to packed 32-bit integers. - _mm256_cvtps_pd⚠(x86 or x86-64) and
avx
Converts packed single-precision (32-bit) floating-point elements ina
to packed double-precision (64-bit) floating-point elements. - _mm256_cvtps_ph⚠(x86 or x86-64) and
f16c
Converts the 8 x 32-bit float values in the 256-bit vectora
into 8 x 16-bit half-precision float values stored in a 128-bit wide vector. - _mm256_cvtsd_f64⚠(x86 or x86-64) and
avx2
Returns the first element of the input vector of[4 x double]
. - _mm256_cvtsi256_si32⚠(x86 or x86-64) and
avx2
Returns the first element of the input vector of[8 x i32]
. - _mm256_cvtss_f32⚠(x86 or x86-64) and
avx
Returns the first element of the input vector of[8 x float]
. - _mm256_cvttpd_epi32⚠(x86 or x86-64) and
avx
Converts packed double-precision (64-bit) floating-point elements ina
to packed 32-bit integers with truncation. - _mm256_cvttps_epi32⚠(x86 or x86-64) and
avx
Converts packed single-precision (32-bit) floating-point elements ina
to packed 32-bit integers with truncation. - _mm256_div_pd⚠(x86 or x86-64) and
avx
Computes the division of each of the 4 packed 64-bit floating-point elements ina
by the corresponding packed elements inb
. - _mm256_div_ps⚠(x86 or x86-64) and
avx
Computes the division of each of the 8 packed 32-bit floating-point elements ina
by the corresponding packed elements inb
. - _mm256_dp_ps⚠(x86 or x86-64) and
avx
Conditionally multiplies the packed single-precision (32-bit) floating-point elements ina
andb
using the high 4 bits inimm8
, sum the four products, and conditionally return the sum using the low 4 bits ofimm8
. - _mm256_extract_epi8⚠(x86 or x86-64) and
avx2
Extracts an 8-bit integer froma
, selected withINDEX
. Returns a 32-bit integer containing the zero-extended integer data. - _mm256_extract_epi16⚠(x86 or x86-64) and
avx2
Extracts a 16-bit integer froma
, selected withINDEX
. Returns a 32-bit integer containing the zero-extended integer data. - _mm256_extract_epi32⚠(x86 or x86-64) and
avx2
Extracts a 32-bit integer froma
, selected withINDEX
. - _mm256_extractf128_pd⚠(x86 or x86-64) and
avx
Extracts 128 bits (composed of 2 packed double-precision (64-bit) floating-point elements) froma
, selected withimm8
. - _mm256_extractf128_ps⚠(x86 or x86-64) and
avx
Extracts 128 bits (composed of 4 packed single-precision (32-bit) floating-point elements) froma
, selected withimm8
. - _mm256_extractf128_si256⚠(x86 or x86-64) and
avx
Extracts 128 bits (composed of integer data) froma
, selected withimm8
. - _mm256_extracti128_si256⚠(x86 or x86-64) and
avx2
Extracts 128 bits (of integer data) froma
selected withIMM1
. - _mm256_floor_pd⚠(x86 or x86-64) and
avx
Rounds packed double-precision (64-bit) floating point elements ina
toward negative infinity. - _mm256_floor_ps⚠(x86 or x86-64) and
avx
Rounds packed single-precision (32-bit) floating point elements ina
toward negative infinity. - _mm256_fmadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and add the intermediate result to packed elements inc
. - _mm256_fmadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and add the intermediate result to packed elements inc
. - _mm256_fmaddsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - _mm256_fmaddsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - _mm256_fmsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the intermediate result. - _mm256_fmsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the intermediate result. - _mm256_fmsubadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - _mm256_fmsubadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - _mm256_fnmadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and add the negated intermediate result to packed elements inc
. - _mm256_fnmadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and add the negated intermediate result to packed elements inc
. - _mm256_fnmsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the negated intermediate result. - _mm256_fnmsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the negated intermediate result. - _mm256_hadd_epi16⚠(x86 or x86-64) and
avx2
Horizontally adds adjacent pairs of 16-bit integers ina
andb
. - _mm256_hadd_epi32⚠(x86 or x86-64) and
avx2
Horizontally adds adjacent pairs of 32-bit integers ina
andb
. - _mm256_hadd_pd⚠(x86 or x86-64) and
avx
Horizontal addition of adjacent pairs in the two packed vectors of 4 64-bit floating pointsa
andb
. In the result, sums of elements froma
are returned in even locations, while sums of elements fromb
are returned in odd locations. - _mm256_hadd_ps⚠(x86 or x86-64) and
avx
Horizontal addition of adjacent pairs in the two packed vectors of 8 32-bit floating pointsa
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. - _mm256_hadds_epi16⚠(x86 or x86-64) and
avx2
Horizontally adds adjacent pairs of 16-bit integers ina
andb
using saturation. - _mm256_hsub_epi16⚠(x86 or x86-64) and
avx2
Horizontally subtract adjacent pairs of 16-bit integers ina
andb
. - _mm256_hsub_epi32⚠(x86 or x86-64) and
avx2
Horizontally subtract adjacent pairs of 32-bit integers ina
andb
. - _mm256_hsub_pd⚠(x86 or x86-64) and
avx
Horizontal subtraction of adjacent pairs in the two packed vectors of 4 64-bit floating pointsa
andb
. In the result, sums of elements froma
are returned in even locations, while sums of elements fromb
are returned in odd locations. - _mm256_hsub_ps⚠(x86 or x86-64) and
avx
Horizontal subtraction of adjacent pairs in the two packed vectors of 8 32-bit floating pointsa
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. - _mm256_hsubs_epi16⚠(x86 or x86-64) and
avx2
Horizontally subtract adjacent pairs of 16-bit integers ina
andb
using saturation. - _mm256_i32gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i32gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i32gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i32gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i64gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i64gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i64gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_i64gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm256_insert_epi8⚠(x86 or x86-64) and
avx
Copiesa
to result, and inserts the 8-bit integeri
into result at the location specified byindex
. - _mm256_insert_epi16⚠(x86 or x86-64) and
avx
Copiesa
to result, and inserts the 16-bit integeri
into result at the location specified byindex
. - _mm256_insert_epi32⚠(x86 or x86-64) and
avx
Copiesa
to result, and inserts the 32-bit integeri
into result at the location specified byindex
. - _mm256_insertf128_pd⚠(x86 or x86-64) and
avx
Copiesa
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
. - _mm256_insertf128_ps⚠(x86 or x86-64) and
avx
Copiesa
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
. - _mm256_insertf128_si256⚠(x86 or x86-64) and
avx
Copiesa
to result, then inserts 128 bits fromb
into result at the location specified byimm8
. - _mm256_inserti128_si256⚠(x86 or x86-64) and
avx2
Copiesa
todst
, then insert 128 bits (of integer data) fromb
at the location specified byIMM1
. - _mm256_lddqu_si256⚠(x86 or x86-64) and
avx
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. - _mm256_load_pd⚠(x86 or x86-64) and
avx
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. - _mm256_load_ps⚠(x86 or x86-64) and
avx
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. - _mm256_load_si256⚠(x86 or x86-64) and
avx
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. - _mm256_loadu2_m128⚠(x86 or x86-64) and
avx,sse
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. - _mm256_loadu2_m128d⚠(x86 or x86-64) and
avx,sse2
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. - _mm256_loadu2_m128i⚠(x86 or x86-64) and
avx,sse2
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. - _mm256_loadu_pd⚠(x86 or x86-64) and
avx
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. - _mm256_loadu_ps⚠(x86 or x86-64) and
avx
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. - _mm256_loadu_si256⚠(x86 or x86-64) and
avx
Loads 256-bits of integer data from memory into result.mem_addr
does not need to be aligned on any particular boundary. - _mm256_madd_epi16⚠(x86 or x86-64) and
avx2
Multiplies packed signed 16-bit integers ina
andb
, producing intermediate signed 32-bit integers. Horizontally add adjacent pairs of intermediate 32-bit integers. - _mm256_maddubs_epi16⚠(x86 or x86-64) and
avx2
Vertically multiplies each unsigned 8-bit integer froma
with the corresponding signed 8-bit integer fromb
, producing intermediate signed 16-bit integers. Horizontally add adjacent pairs of intermediate signed 16-bit integers - _mm256_mask_i32gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i32gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i32gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i32gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i64gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i64gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i64gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_mask_i64gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm256_maskload_epi32⚠(x86 or x86-64) and
avx2
Loads packed 32-bit integers from memory pointed bymem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - _mm256_maskload_epi64⚠(x86 or x86-64) and
avx2
Loads packed 64-bit integers from memory pointed bymem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - _mm256_maskload_pd⚠(x86 or x86-64) and
avx
Loads packed double-precision (64-bit) floating-point elements from memory into result usingmask
(elements are zeroed out when the high bit of the corresponding element is not set). - _mm256_maskload_ps⚠(x86 or x86-64) and
avx
Loads packed single-precision (32-bit) floating-point elements from memory into result usingmask
(elements are zeroed out when the high bit of the corresponding element is not set). - _mm256_maskstore_epi32⚠(x86 or x86-64) and
avx2
Stores packed 32-bit integers froma
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - _mm256_maskstore_epi64⚠(x86 or x86-64) and
avx2
Stores packed 64-bit integers froma
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - _mm256_maskstore_pd⚠(x86 or x86-64) and
avx
Stores packed double-precision (64-bit) floating-point elements froma
into memory usingmask
. - _mm256_maskstore_ps⚠(x86 or x86-64) and
avx
Stores packed single-precision (32-bit) floating-point elements froma
into memory usingmask
. - _mm256_max_epi8⚠(x86 or x86-64) and
avx2
Compares packed 8-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_epi16⚠(x86 or x86-64) and
avx2
Compares packed 16-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_epi32⚠(x86 or x86-64) and
avx2
Compares packed 32-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_epu8⚠(x86 or x86-64) and
avx2
Compares packed unsigned 8-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_epu16⚠(x86 or x86-64) and
avx2
Compares packed unsigned 16-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_epu32⚠(x86 or x86-64) and
avx2
Compares packed unsigned 32-bit integers ina
andb
, and returns the packed maximum values. - _mm256_max_pd⚠(x86 or x86-64) and
avx
Compares packed double-precision (64-bit) floating-point elements ina
andb
, and returns packed maximum values - _mm256_max_ps⚠(x86 or x86-64) and
avx
Compares packed single-precision (32-bit) floating-point elements ina
andb
, and returns packed maximum values - _mm256_min_epi8⚠(x86 or x86-64) and
avx2
Compares packed 8-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_epi16⚠(x86 or x86-64) and
avx2
Compares packed 16-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_epi32⚠(x86 or x86-64) and
avx2
Compares packed 32-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_epu8⚠(x86 or x86-64) and
avx2
Compares packed unsigned 8-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_epu16⚠(x86 or x86-64) and
avx2
Compares packed unsigned 16-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_epu32⚠(x86 or x86-64) and
avx2
Compares packed unsigned 32-bit integers ina
andb
, and returns the packed minimum values. - _mm256_min_pd⚠(x86 or x86-64) and
avx
Compares packed double-precision (64-bit) floating-point elements ina
andb
, and returns packed minimum values - _mm256_min_ps⚠(x86 or x86-64) and
avx
Compares packed single-precision (32-bit) floating-point elements ina
andb
, and returns packed minimum values - _mm256_movedup_pd⚠(x86 or x86-64) and
avx
Duplicate even-indexed double-precision (64-bit) floating-point elements froma
, and returns the results. - _mm256_movehdup_ps⚠(x86 or x86-64) and
avx
Duplicate odd-indexed single-precision (32-bit) floating-point elements froma
, and returns the results. - _mm256_moveldup_ps⚠(x86 or x86-64) and
avx
Duplicate even-indexed single-precision (32-bit) floating-point elements froma
, and returns the results. - _mm256_movemask_epi8⚠(x86 or x86-64) and
avx2
Creates mask from the most significant bit of each 8-bit element ina
, return the result. - _mm256_movemask_pd⚠(x86 or x86-64) and
avx
Sets each bit of the returned mask based on the most significant bit of the corresponding packed double-precision (64-bit) floating-point element ina
. - _mm256_movemask_ps⚠(x86 or x86-64) and
avx
Sets each bit of the returned mask based on the most significant bit of the corresponding packed single-precision (32-bit) floating-point element ina
. - _mm256_mpsadbw_epu8⚠(x86 or x86-64) and
avx2
Computes the sum of absolute differences (SADs) of quadruplets of unsigned 8-bit integers ina
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
. - _mm256_mul_epi32⚠(x86 or x86-64) and
avx2
Multiplies the low 32-bit integers from each packed 64-bit element ina
andb
- _mm256_mul_epu32⚠(x86 or x86-64) and
avx2
Multiplies the low unsigned 32-bit integers from each packed 64-bit element ina
andb
- _mm256_mul_pd⚠(x86 or x86-64) and
avx
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
. - _mm256_mul_ps⚠(x86 or x86-64) and
avx
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
. - _mm256_mulhi_epi16⚠(x86 or x86-64) and
avx2
Multiplies the packed 16-bit integers ina
andb
, producing intermediate 32-bit integers and returning the high 16 bits of the intermediate integers. - _mm256_mulhi_epu16⚠(x86 or x86-64) and
avx2
Multiplies the packed unsigned 16-bit integers ina
andb
, producing intermediate 32-bit integers and returning the high 16 bits of the intermediate integers. - _mm256_mulhrs_epi16⚠(x86 or x86-64) and
avx2
Multiplies packed 16-bit integers ina
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]
. - _mm256_mullo_epi16⚠(x86 or x86-64) and
avx2
Multiplies the packed 16-bit integers ina
andb
, producing intermediate 32-bit integers, and returns the low 16 bits of the intermediate integers - _mm256_mullo_epi32⚠(x86 or x86-64) and
avx2
Multiplies the packed 32-bit integers ina
andb
, producing intermediate 64-bit integers, and returns the low 32 bits of the intermediate integers - _mm256_or_pd⚠(x86 or x86-64) and
avx
Computes the bitwise OR packed double-precision (64-bit) floating-point elements ina
andb
. - _mm256_or_ps⚠(x86 or x86-64) and
avx
Computes the bitwise OR packed single-precision (32-bit) floating-point elements ina
andb
. - _mm256_or_si256⚠(x86 or x86-64) and
avx2
Computes the bitwise OR of 256 bits (representing integer data) ina
andb
- _mm256_packs_epi16⚠(x86 or x86-64) and
avx2
Converts packed 16-bit integers froma
andb
to packed 8-bit integers using signed saturation - _mm256_packs_epi32⚠(x86 or x86-64) and
avx2
Converts packed 32-bit integers froma
andb
to packed 16-bit integers using signed saturation - _mm256_packus_epi16⚠(x86 or x86-64) and
avx2
Converts packed 16-bit integers froma
andb
to packed 8-bit integers using unsigned saturation - _mm256_packus_epi32⚠(x86 or x86-64) and
avx2
Converts packed 32-bit integers froma
andb
to packed 16-bit integers using unsigned saturation - _mm256_permute2f128_pd⚠(x86 or x86-64) and
avx
Shuffles 256 bits (composed of 4 packed double-precision (64-bit) floating-point elements) selected byimm8
froma
andb
. - _mm256_permute2f128_ps⚠(x86 or x86-64) and
avx
Shuffles 256 bits (composed of 8 packed single-precision (32-bit) floating-point elements) selected byimm8
froma
andb
. - _mm256_permute2f128_si256⚠(x86 or x86-64) and
avx
Shuffles 128-bits (composed of integer data) selected byimm8
froma
andb
. - _mm256_permute2x128_si256⚠(x86 or x86-64) and
avx2
Shuffles 128-bits of integer data selected byimm8
froma
andb
. - _mm256_permute4x64_epi64⚠(x86 or x86-64) and
avx2
Permutes 64-bit integers froma
using control maskimm8
. - _mm256_permute4x64_pd⚠(x86 or x86-64) and
avx2
Shuffles 64-bit floating-point elements ina
across lanes using the control inimm8
. - _mm256_permute_pd⚠(x86 or x86-64) and
avx
Shuffles double-precision (64-bit) floating-point elements ina
within 128-bit lanes using the control inimm8
. - _mm256_permute_ps⚠(x86 or x86-64) and
avx
Shuffles single-precision (32-bit) floating-point elements ina
within 128-bit lanes using the control inimm8
. - _mm256_permutevar8x32_epi32⚠(x86 or x86-64) and
avx2
Permutes packed 32-bit integers froma
according to the content ofb
. - _mm256_permutevar8x32_ps⚠(x86 or x86-64) and
avx2
Shuffles eight 32-bit floating-point elements ina
across lanes using the corresponding 32-bit integer index inidx
. - _mm256_permutevar_pd⚠(x86 or x86-64) and
avx
Shuffles double-precision (64-bit) floating-point elements ina
within 256-bit lanes using the control inb
. - _mm256_permutevar_ps⚠(x86 or x86-64) and
avx
Shuffles single-precision (32-bit) floating-point elements ina
within 128-bit lanes using the control inb
. - _mm256_rcp_ps⚠(x86 or x86-64) and
avx
Computes the approximate reciprocal of packed single-precision (32-bit) floating-point elements ina
, and returns the results. The maximum relative error for this approximation is less than 1.5*2^-12. - _mm256_round_pd⚠(x86 or x86-64) and
avx
Rounds packed double-precision (64-bit) floating point elements ina
according to the flagROUNDING
. The value ofROUNDING
may be as follows: - _mm256_round_ps⚠(x86 or x86-64) and
avx
Rounds packed single-precision (32-bit) floating point elements ina
according to the flagROUNDING
. The value ofROUNDING
may be as follows: - _mm256_rsqrt_ps⚠(x86 or x86-64) and
avx
Computes the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements ina
, and returns the results. The maximum relative error for this approximation is less than 1.5*2^-12. - _mm256_sad_epu8⚠(x86 or x86-64) and
avx2
Computes the absolute differences of packed unsigned 8-bit integers ina
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 - _mm256_set1_epi8⚠(x86 or x86-64) and
avx
Broadcasts 8-bit integera
to all elements of returned vector. This intrinsic may generate thevpbroadcastb
. - _mm256_set1_epi16⚠(x86 or x86-64) and
avx
Broadcasts 16-bit integera
to all elements of returned vector. This intrinsic may generate thevpbroadcastw
. - _mm256_set1_epi32⚠(x86 or x86-64) and
avx
Broadcasts 32-bit integera
to all elements of returned vector. This intrinsic may generate thevpbroadcastd
. - _mm256_set1_epi64x⚠(x86 or x86-64) and
avx
Broadcasts 64-bit integera
to all elements of returned vector. This intrinsic may generate thevpbroadcastq
. - _mm256_set1_pd⚠(x86 or x86-64) and
avx
Broadcasts double-precision (64-bit) floating-point valuea
to all elements of returned vector. - _mm256_set1_ps⚠(x86 or x86-64) and
avx
Broadcasts single-precision (32-bit) floating-point valuea
to all elements of returned vector. - _mm256_set_epi8⚠(x86 or x86-64) and
avx
Sets packed 8-bit integers in returned vector with the supplied values. - _mm256_set_epi16⚠(x86 or x86-64) and
avx
Sets packed 16-bit integers in returned vector with the supplied values. - _mm256_set_epi32⚠(x86 or x86-64) and
avx
Sets packed 32-bit integers in returned vector with the supplied values. - _mm256_set_epi64x⚠(x86 or x86-64) and
avx
Sets packed 64-bit integers in returned vector with the supplied values. - _mm256_set_m128⚠(x86 or x86-64) and
avx
Sets packed __m256 returned vector with the supplied values. - _mm256_set_m128d⚠(x86 or x86-64) and
avx
Sets packed __m256d returned vector with the supplied values. - _mm256_set_m128i⚠(x86 or x86-64) and
avx
Sets packed __m256i returned vector with the supplied values. - _mm256_set_pd⚠(x86 or x86-64) and
avx
Sets packed double-precision (64-bit) floating-point elements in returned vector with the supplied values. - _mm256_set_ps⚠(x86 or x86-64) and
avx
Sets packed single-precision (32-bit) floating-point elements in returned vector with the supplied values. - _mm256_setr_epi8⚠(x86 or x86-64) and
avx
Sets packed 8-bit integers in returned vector with the supplied values in reverse order. - _mm256_setr_epi16⚠(x86 or x86-64) and
avx
Sets packed 16-bit integers in returned vector with the supplied values in reverse order. - _mm256_setr_epi32⚠(x86 or x86-64) and
avx
Sets packed 32-bit integers in returned vector with the supplied values in reverse order. - _mm256_setr_epi64x⚠(x86 or x86-64) and
avx
Sets packed 64-bit integers in returned vector with the supplied values in reverse order. - _mm256_setr_m128⚠(x86 or x86-64) and
avx
Sets packed __m256 returned vector with the supplied values. - _mm256_setr_m128d⚠(x86 or x86-64) and
avx
Sets packed __m256d returned vector with the supplied values. - _mm256_setr_m128i⚠(x86 or x86-64) and
avx
Sets packed __m256i returned vector with the supplied values. - _mm256_setr_pd⚠(x86 or x86-64) and
avx
Sets packed double-precision (64-bit) floating-point elements in returned vector with the supplied values in reverse order. - _mm256_setr_ps⚠(x86 or x86-64) and
avx
Sets packed single-precision (32-bit) floating-point elements in returned vector with the supplied values in reverse order. - _mm256_setzero_pd⚠(x86 or x86-64) and
avx
Returns vector of type __m256d with all elements set to zero. - _mm256_setzero_ps⚠(x86 or x86-64) and
avx
Returns vector of type __m256 with all elements set to zero. - _mm256_setzero_si256⚠(x86 or x86-64) and
avx
Returns vector of type __m256i with all elements set to zero. - _mm256_shuffle_epi8⚠(x86 or x86-64) and
avx2
Shuffles bytes froma
according to the content ofb
. - _mm256_shuffle_epi32⚠(x86 or x86-64) and
avx2
Shuffles 32-bit integers in 128-bit lanes ofa
using the control inimm8
. - _mm256_shuffle_pd⚠(x86 or x86-64) and
avx
Shuffles double-precision (64-bit) floating-point elements within 128-bit lanes using the control inimm8
. - _mm256_shuffle_ps⚠(x86 or x86-64) and
avx
Shuffles single-precision (32-bit) floating-point elements ina
within 128-bit lanes using the control inimm8
. - _mm256_shufflehi_epi16⚠(x86 or x86-64) and
avx2
Shuffles 16-bit integers in the high 64 bits of 128-bit lanes ofa
using the control inimm8
. The low 64 bits of 128-bit lanes ofa
are copied to the output. - _mm256_shufflelo_epi16⚠(x86 or x86-64) and
avx2
Shuffles 16-bit integers in the low 64 bits of 128-bit lanes ofa
using the control inimm8
. The high 64 bits of 128-bit lanes ofa
are copied to the output. - _mm256_sign_epi8⚠(x86 or x86-64) and
avx2
Negates packed 8-bit integers ina
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. - _mm256_sign_epi16⚠(x86 or x86-64) and
avx2
Negates packed 16-bit integers ina
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. - _mm256_sign_epi32⚠(x86 or x86-64) and
avx2
Negates packed 32-bit integers ina
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. - _mm256_sll_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
left bycount
while shifting in zeros, and returns the result - _mm256_sll_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
left bycount
while shifting in zeros, and returns the result - _mm256_sll_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
left bycount
while shifting in zeros, and returns the result - _mm256_slli_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
left byIMM8
while shifting in zeros, return the results; - _mm256_slli_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
left byIMM8
while shifting in zeros, return the results; - _mm256_slli_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
left byIMM8
while shifting in zeros, return the results; - _mm256_slli_si256⚠(x86 or x86-64) and
avx2
Shifts 128-bit lanes ina
left byimm8
bytes while shifting in zeros. - _mm256_sllv_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - _mm256_sllv_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - _mm256_sqrt_pd⚠(x86 or x86-64) and
avx
Returns the square root of packed double-precision (64-bit) floating point elements ina
. - _mm256_sqrt_ps⚠(x86 or x86-64) and
avx
Returns the square root of packed single-precision (32-bit) floating point elements ina
. - _mm256_sra_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
right bycount
while shifting in sign bits. - _mm256_sra_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right bycount
while shifting in sign bits. - _mm256_srai_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
right byIMM8
while shifting in sign bits. - _mm256_srai_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right byIMM8
while shifting in sign bits. - _mm256_srav_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in sign bits. - _mm256_srl_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
right bycount
while shifting in zeros. - _mm256_srl_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right bycount
while shifting in zeros. - _mm256_srl_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
right bycount
while shifting in zeros. - _mm256_srli_epi16⚠(x86 or x86-64) and
avx2
Shifts packed 16-bit integers ina
right byIMM8
while shifting in zeros - _mm256_srli_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right byIMM8
while shifting in zeros - _mm256_srli_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
right byIMM8
while shifting in zeros - _mm256_srli_si256⚠(x86 or x86-64) and
avx2
Shifts 128-bit lanes ina
right byimm8
bytes while shifting in zeros. - _mm256_srlv_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in zeros, - _mm256_srlv_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in zeros, - _mm256_store_pd⚠(x86 or x86-64) and
avx
Stores 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) froma
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_store_ps⚠(x86 or x86-64) and
avx
Stores 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) froma
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_store_si256⚠(x86 or x86-64) and
avx
Stores 256-bits of integer data froma
into memory.mem_addr
must be aligned on a 32-byte boundary or a general-protection exception may be generated. - _mm256_storeu2_m128⚠(x86 or x86-64) and
avx,sse
Stores the high and low 128-bit halves (each composed of 4 packed single-precision (32-bit) floating-point elements) froma
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - _mm256_storeu2_m128d⚠(x86 or x86-64) and
avx,sse2
Stores the high and low 128-bit halves (each composed of 2 packed double-precision (64-bit) floating-point elements) froma
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - _mm256_storeu2_m128i⚠(x86 or x86-64) and
avx,sse2
Stores the high and low 128-bit halves (each composed of integer data) froma
into memory two different 128-bit locations.hiaddr
andloaddr
do not need to be aligned on any particular boundary. - _mm256_storeu_pd⚠(x86 or x86-64) and
avx
Stores 256-bits (composed of 4 packed double-precision (64-bit) floating-point elements) froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm256_storeu_ps⚠(x86 or x86-64) and
avx
Stores 256-bits (composed of 8 packed single-precision (32-bit) floating-point elements) froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm256_storeu_si256⚠(x86 or x86-64) and
avx
Stores 256-bits of integer data froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm256_stream_pd⚠(x86 or x86-64) and
avx
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). - _mm256_stream_ps⚠(x86 or x86-64) and
avx
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). - _mm256_stream_si256⚠(x86 or x86-64) and
avx
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) - _mm256_sub_epi8⚠(x86 or x86-64) and
avx2
Subtract packed 8-bit integers inb
from packed 8-bit integers ina
- _mm256_sub_epi16⚠(x86 or x86-64) and
avx2
Subtract packed 16-bit integers inb
from packed 16-bit integers ina
- _mm256_sub_epi32⚠(x86 or x86-64) and
avx2
Subtract packed 32-bit integers inb
from packed 32-bit integers ina
- _mm256_sub_epi64⚠(x86 or x86-64) and
avx2
Subtract packed 64-bit integers inb
from packed 64-bit integers ina
- _mm256_sub_pd⚠(x86 or x86-64) and
avx
Subtracts packed double-precision (64-bit) floating-point elements inb
from packed elements ina
. - _mm256_sub_ps⚠(x86 or x86-64) and
avx
Subtracts packed single-precision (32-bit) floating-point elements inb
from packed elements ina
. - _mm256_subs_epi8⚠(x86 or x86-64) and
avx2
Subtract packed 8-bit integers inb
from packed 8-bit integers ina
using saturation. - _mm256_subs_epi16⚠(x86 or x86-64) and
avx2
Subtract packed 16-bit integers inb
from packed 16-bit integers ina
using saturation. - _mm256_subs_epu8⚠(x86 or x86-64) and
avx2
Subtract packed unsigned 8-bit integers inb
from packed 8-bit integers ina
using saturation. - _mm256_subs_epu16⚠(x86 or x86-64) and
avx2
Subtract packed unsigned 16-bit integers inb
from packed 16-bit integers ina
using saturation. - _mm256_testc_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm256_testc_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm256_testc_si256⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing integer data) ina
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. - _mm256_testnzc_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm256_testnzc_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm256_testnzc_si256⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing integer data) ina
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. - _mm256_testz_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm256_testz_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm256_testz_si256⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 256 bits (representing integer data) ina
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. - _mm256_undefined_pd⚠(x86 or x86-64) and
avx
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
. - _mm256_undefined_ps⚠(x86 or x86-64) and
avx
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
. - _mm256_undefined_si256⚠(x86 or x86-64) and
avx
Returns vector of type __m256i with with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm256_unpackhi_epi8⚠(x86 or x86-64) and
avx2
Unpacks and interleave 8-bit integers from the high half of each 128-bit lane ina
andb
. - _mm256_unpackhi_epi16⚠(x86 or x86-64) and
avx2
Unpacks and interleave 16-bit integers from the high half of each 128-bit lane ofa
andb
. - _mm256_unpackhi_epi32⚠(x86 or x86-64) and
avx2
Unpacks and interleave 32-bit integers from the high half of each 128-bit lane ofa
andb
. - _mm256_unpackhi_epi64⚠(x86 or x86-64) and
avx2
Unpacks and interleave 64-bit integers from the high half of each 128-bit lane ofa
andb
. - _mm256_unpackhi_pd⚠(x86 or x86-64) and
avx
Unpacks and interleave double-precision (64-bit) floating-point elements from the high half of each 128-bit lane ina
andb
. - _mm256_unpackhi_ps⚠(x86 or x86-64) and
avx
Unpacks and interleave single-precision (32-bit) floating-point elements from the high half of each 128-bit lane ina
andb
. - _mm256_unpacklo_epi8⚠(x86 or x86-64) and
avx2
Unpacks and interleave 8-bit integers from the low half of each 128-bit lane ofa
andb
. - _mm256_unpacklo_epi16⚠(x86 or x86-64) and
avx2
Unpacks and interleave 16-bit integers from the low half of each 128-bit lane ofa
andb
. - _mm256_unpacklo_epi32⚠(x86 or x86-64) and
avx2
Unpacks and interleave 32-bit integers from the low half of each 128-bit lane ofa
andb
. - _mm256_unpacklo_epi64⚠(x86 or x86-64) and
avx2
Unpacks and interleave 64-bit integers from the low half of each 128-bit lane ofa
andb
. - _mm256_unpacklo_pd⚠(x86 or x86-64) and
avx
Unpacks and interleave double-precision (64-bit) floating-point elements from the low half of each 128-bit lane ina
andb
. - _mm256_unpacklo_ps⚠(x86 or x86-64) and
avx
Unpacks and interleave single-precision (32-bit) floating-point elements from the low half of each 128-bit lane ina
andb
. - _mm256_xor_pd⚠(x86 or x86-64) and
avx
Computes the bitwise XOR of packed double-precision (64-bit) floating-point elements ina
andb
. - _mm256_xor_ps⚠(x86 or x86-64) and
avx
Computes the bitwise XOR of packed single-precision (32-bit) floating-point elements ina
andb
. - _mm256_xor_si256⚠(x86 or x86-64) and
avx2
Computes the bitwise XOR of 256 bits (representing integer data) ina
andb
- _mm256_zeroall⚠(x86 or x86-64) and
avx
Zeroes the contents of all XMM or YMM registers. - _mm256_zeroupper⚠(x86 or x86-64) and
avx
Zeroes the upper 128 bits of all YMM registers; the lower 128-bits of the registers are unmodified. - _mm256_zextpd128_pd256⚠(x86 or x86-64) and
avx,sse2
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. - _mm256_zextps128_ps256⚠(x86 or x86-64) and
avx,sse
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. - _mm256_zextsi128_si256⚠(x86 or x86-64) and
avx,sse2
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. - _mm_abs_epi8⚠(x86 or x86-64) and
ssse3
Computes the absolute value of packed 8-bit signed integers ina
and return the unsigned results. - _mm_abs_epi16⚠(x86 or x86-64) and
ssse3
Computes the absolute value of each of the packed 16-bit signed integers ina
and return the 16-bit unsigned integer - _mm_abs_epi32⚠(x86 or x86-64) and
ssse3
Computes the absolute value of each of the packed 32-bit signed integers ina
and return the 32-bit unsigned integer - _mm_add_epi8⚠(x86 or x86-64) and
sse2
Adds packed 8-bit integers ina
andb
. - _mm_add_epi16⚠(x86 or x86-64) and
sse2
Adds packed 16-bit integers ina
andb
. - _mm_add_epi32⚠(x86 or x86-64) and
sse2
Adds packed 32-bit integers ina
andb
. - _mm_add_epi64⚠(x86 or x86-64) and
sse2
Adds packed 64-bit integers ina
andb
. - _mm_add_pd⚠(x86 or x86-64) and
sse2
Adds packed double-precision (64-bit) floating-point elements ina
andb
. - _mm_add_ps⚠(x86 or x86-64) and
sse
Adds __m128 vectors. - _mm_add_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the sum of the low elements ofa
andb
. - _mm_add_ss⚠(x86 or x86-64) and
sse
Adds the first component ofa
andb
, the other components are copied froma
. - _mm_adds_epi8⚠(x86 or x86-64) and
sse2
Adds packed 8-bit integers ina
andb
using saturation. - _mm_adds_epi16⚠(x86 or x86-64) and
sse2
Adds packed 16-bit integers ina
andb
using saturation. - _mm_adds_epu8⚠(x86 or x86-64) and
sse2
Adds packed unsigned 8-bit integers ina
andb
using saturation. - _mm_adds_epu16⚠(x86 or x86-64) and
sse2
Adds packed unsigned 16-bit integers ina
andb
using saturation. - _mm_addsub_pd⚠(x86 or x86-64) and
sse3
Alternatively add and subtract packed double-precision (64-bit) floating-point elements ina
to/from packed elements inb
. - _mm_addsub_ps⚠(x86 or x86-64) and
sse3
Alternatively add and subtract packed single-precision (32-bit) floating-point elements ina
to/from packed elements inb
. - _mm_aesdec_si128⚠(x86 or x86-64) and
aes
Performs one round of an AES decryption flow on data (state) ina
. - _mm_aesdeclast_si128⚠(x86 or x86-64) and
aes
Performs the last round of an AES decryption flow on data (state) ina
. - _mm_aesenc_si128⚠(x86 or x86-64) and
aes
Performs one round of an AES encryption flow on data (state) ina
. - _mm_aesenclast_si128⚠(x86 or x86-64) and
aes
Performs the last round of an AES encryption flow on data (state) ina
. - _mm_aesimc_si128⚠(x86 or x86-64) and
aes
Performs theInvMixColumns
transformation ona
. - _mm_aeskeygenassist_si128⚠(x86 or x86-64) and
aes
Assist in expanding the AES cipher key. - _mm_alignr_epi8⚠(x86 or x86-64) and
ssse3
Concatenate 16-byte blocks ina
andb
into a 32-byte temporary result, shift the result right byn
bytes, and returns the low 16 bytes. - _mm_and_pd⚠(x86 or x86-64) and
sse2
Computes the bitwise AND of packed double-precision (64-bit) floating-point elements ina
andb
. - _mm_and_ps⚠(x86 or x86-64) and
sse
Bitwise AND of packed single-precision (32-bit) floating-point elements. - _mm_and_si128⚠(x86 or x86-64) and
sse2
Computes the bitwise AND of 128 bits (representing integer data) ina
andb
. - _mm_andnot_pd⚠(x86 or x86-64) and
sse2
Computes the bitwise NOT ofa
and then AND withb
. - _mm_andnot_ps⚠(x86 or x86-64) and
sse
Bitwise AND-NOT of packed single-precision (32-bit) floating-point elements. - _mm_andnot_si128⚠(x86 or x86-64) and
sse2
Computes the bitwise NOT of 128 bits (representing integer data) ina
and then AND withb
. - _mm_avg_epu8⚠(x86 or x86-64) and
sse2
Averages packed unsigned 8-bit integers ina
andb
. - _mm_avg_epu16⚠(x86 or x86-64) and
sse2
Averages packed unsigned 16-bit integers ina
andb
. - _mm_blend_epi16⚠(x86 or x86-64) and
sse4.1
Blend packed 16-bit integers froma
andb
using the maskIMM8
. - _mm_blend_epi32⚠(x86 or x86-64) and
avx2
Blends packed 32-bit integers froma
andb
using control maskIMM4
. - _mm_blend_pd⚠(x86 or x86-64) and
sse4.1
Blend packed double-precision (64-bit) floating-point elements froma
andb
using control maskIMM2
- _mm_blend_ps⚠(x86 or x86-64) and
sse4.1
Blend packed single-precision (32-bit) floating-point elements froma
andb
using maskIMM4
- _mm_blendv_epi8⚠(x86 or x86-64) and
sse4.1
Blend packed 8-bit integers froma
andb
usingmask
- _mm_blendv_pd⚠(x86 or x86-64) and
sse4.1
Blend packed double-precision (64-bit) floating-point elements froma
andb
usingmask
- _mm_blendv_ps⚠(x86 or x86-64) and
sse4.1
Blend packed single-precision (32-bit) floating-point elements froma
andb
usingmask
- _mm_broadcast_ss⚠(x86 or x86-64) and
avx
Broadcasts a single-precision (32-bit) floating-point element from memory to all elements of the returned vector. - _mm_broadcastb_epi8⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 8-bit integer froma
to all elements of the 128-bit returned value. - _mm_broadcastd_epi32⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 32-bit integer froma
to all elements of the 128-bit returned value. - _mm_broadcastq_epi64⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 64-bit integer froma
to all elements of the 128-bit returned value. - _mm_broadcastsd_pd⚠(x86 or x86-64) and
avx2
Broadcasts the low double-precision (64-bit) floating-point element froma
to all elements of the 128-bit returned value. - _mm_broadcastss_ps⚠(x86 or x86-64) and
avx2
Broadcasts the low single-precision (32-bit) floating-point element froma
to all elements of the 128-bit returned value. - _mm_broadcastw_epi16⚠(x86 or x86-64) and
avx2
Broadcasts the low packed 16-bit integer from a to all elements of the 128-bit returned value - _mm_bslli_si128⚠(x86 or x86-64) and
sse2
Shiftsa
left byIMM8
bytes while shifting in zeros. - _mm_bsrli_si128⚠(x86 or x86-64) and
sse2
Shiftsa
right byIMM8
bytes while shifting in zeros. - _mm_castpd_ps⚠(x86 or x86-64) and
sse2
Casts a 128-bit floating-point vector of[2 x double]
into a 128-bit floating-point vector of[4 x float]
. - _mm_castpd_si128⚠(x86 or x86-64) and
sse2
Casts a 128-bit floating-point vector of[2 x double]
into a 128-bit integer vector. - _mm_castps_pd⚠(x86 or x86-64) and
sse2
Casts a 128-bit floating-point vector of[4 x float]
into a 128-bit floating-point vector of[2 x double]
. - _mm_castps_si128⚠(x86 or x86-64) and
sse2
Casts a 128-bit floating-point vector of[4 x float]
into a 128-bit integer vector. - _mm_castsi128_pd⚠(x86 or x86-64) and
sse2
Casts a 128-bit integer vector into a 128-bit floating-point vector of[2 x double]
. - _mm_castsi128_ps⚠(x86 or x86-64) and
sse2
Casts a 128-bit integer vector into a 128-bit floating-point vector of[4 x float]
. - _mm_ceil_pd⚠(x86 or x86-64) and
sse4.1
Round the packed double-precision (64-bit) floating-point elements ina
up to an integer value, and stores the results as packed double-precision floating-point elements. - _mm_ceil_ps⚠(x86 or x86-64) and
sse4.1
Round the packed single-precision (32-bit) floating-point elements ina
up to an integer value, and stores the results as packed single-precision floating-point elements. - _mm_ceil_sd⚠(x86 or x86-64) and
sse4.1
Round the lower double-precision (64-bit) floating-point element inb
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. - _mm_ceil_ss⚠(x86 or x86-64) and
sse4.1
Round the lower single-precision (32-bit) floating-point element inb
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. - _mm_clflush⚠(x86 or x86-64) and
sse2
Invalidates and flushes the cache line that containsp
from all levels of the cache hierarchy. - _mm_clmulepi64_si128⚠(x86 or x86-64) and
pclmulqdq
Performs a carry-less multiplication of two 64-bit polynomials over the finite field GF(2). - _mm_cmp_pd⚠(x86 or x86-64) and
avx,sse2
Compares packed double-precision (64-bit) floating-point elements ina
andb
based on the comparison operand specified byIMM5
. - _mm_cmp_ps⚠(x86 or x86-64) and
avx,sse
Compares packed single-precision (32-bit) floating-point elements ina
andb
based on the comparison operand specified byIMM5
. - _mm_cmp_sd⚠(x86 or x86-64) and
avx,sse2
Compares the lower double-precision (64-bit) floating-point element ina
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. - _mm_cmp_ss⚠(x86 or x86-64) and
avx,sse
Compares the lower single-precision (32-bit) floating-point element ina
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. - _mm_cmpeq_epi8⚠(x86 or x86-64) and
sse2
Compares packed 8-bit integers ina
andb
for equality. - _mm_cmpeq_epi16⚠(x86 or x86-64) and
sse2
Compares packed 16-bit integers ina
andb
for equality. - _mm_cmpeq_epi32⚠(x86 or x86-64) and
sse2
Compares packed 32-bit integers ina
andb
for equality. - _mm_cmpeq_epi64⚠(x86 or x86-64) and
sse4.1
Compares packed 64-bit integers ina
andb
for equality - _mm_cmpeq_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for equality. - _mm_cmpeq_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input elements were equal, or0
otherwise. - _mm_cmpeq_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the equality comparison of the lower elements ofa
andb
. - _mm_cmpeq_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpestra⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
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. - _mm_cmpestrc⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if the resulting mask was non-zero, and0
otherwise. - _mm_cmpestri⚠(x86 or x86-64) and
sse4.2
Compares packed stringsa
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
. - _mm_cmpestrm⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
andb
with lengthsla
andlb
using the control inIMM8
, and return the generated mask. - _mm_cmpestro⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
andb
with lengthsla
andlb
using the control inIMM8
, and return bit0
of the resulting bit mask. - _mm_cmpestrs⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if any character in a was null, and0
otherwise. - _mm_cmpestrz⚠(x86 or x86-64) and
sse4.2
Compares packed strings ina
andb
with lengthsla
andlb
using the control inIMM8
, and return1
if any character inb
was null, and0
otherwise. - _mm_cmpge_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for greater-than-or-equal. - _mm_cmpge_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpge_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the greater-than-or-equal comparison of the lower elements ofa
andb
. - _mm_cmpge_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpgt_epi8⚠(x86 or x86-64) and
sse2
Compares packed 8-bit integers ina
andb
for greater-than. - _mm_cmpgt_epi16⚠(x86 or x86-64) and
sse2
Compares packed 16-bit integers ina
andb
for greater-than. - _mm_cmpgt_epi32⚠(x86 or x86-64) and
sse2
Compares packed 32-bit integers ina
andb
for greater-than. - _mm_cmpgt_epi64⚠(x86 or x86-64) and
sse4.2
Compares packed 64-bit integers ina
andb
for greater-than, return the results. - _mm_cmpgt_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for greater-than. - _mm_cmpgt_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpgt_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the greater-than comparison of the lower elements ofa
andb
. - _mm_cmpgt_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpistra⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and return1
ifb
did not contain a null character and the resulting mask was zero, and0
otherwise. - _mm_cmpistrc⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and return1
if the resulting mask was non-zero, and0
otherwise. - _mm_cmpistri⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
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. - _mm_cmpistrm⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and return the generated mask. - _mm_cmpistro⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and return bit0
of the resulting bit mask. - _mm_cmpistrs⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and returns1
if any character ina
was null, and0
otherwise. - _mm_cmpistrz⚠(x86 or x86-64) and
sse4.2
Compares packed strings with implicit lengths ina
andb
using the control inIMM8
, and return1
if any character inb
was null. and0
otherwise. - _mm_cmple_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for less-than-or-equal - _mm_cmple_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmple_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the less-than-or-equal comparison of the lower elements ofa
andb
. - _mm_cmple_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmplt_epi8⚠(x86 or x86-64) and
sse2
Compares packed 8-bit integers ina
andb
for less-than. - _mm_cmplt_epi16⚠(x86 or x86-64) and
sse2
Compares packed 16-bit integers ina
andb
for less-than. - _mm_cmplt_epi32⚠(x86 or x86-64) and
sse2
Compares packed 32-bit integers ina
andb
for less-than. - _mm_cmplt_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for less-than. - _mm_cmplt_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmplt_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the less-than comparison of the lower elements ofa
andb
. - _mm_cmplt_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpneq_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for not-equal. - _mm_cmpneq_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
to the corresponding element inb
. The result in the output vector will be0xffffffff
if the input elements are not equal, or0
otherwise. - _mm_cmpneq_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the not-equal comparison of the lower elements ofa
andb
. - _mm_cmpneq_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpnge_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for not-greater-than-or-equal. - _mm_cmpnge_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpnge_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the not-greater-than-or-equal comparison of the lower elements ofa
andb
. - _mm_cmpnge_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpngt_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for not-greater-than. - _mm_cmpngt_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpngt_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the not-greater-than comparison of the lower elements ofa
andb
. - _mm_cmpngt_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpnle_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for not-less-than-or-equal. - _mm_cmpnle_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpnle_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the not-less-than-or-equal comparison of the lower elements ofa
andb
. - _mm_cmpnle_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpnlt_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
for not-less-than. - _mm_cmpnlt_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpnlt_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the not-less-than comparison of the lower elements ofa
andb
. - _mm_cmpnlt_ss⚠(x86 or x86-64) and
sse
Compares the lowestf32
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
. - _mm_cmpord_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
to see if neither isNaN
. - _mm_cmpord_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpord_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the result of comparing both of the lower elements ofa
andb
toNaN
. If neither are equal toNaN
then0xFFFFFFFFFFFFFFFF
is used and0
otherwise. - _mm_cmpord_ss⚠(x86 or x86-64) and
sse
Checks if the lowestf32
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
. - _mm_cmpunord_pd⚠(x86 or x86-64) and
sse2
Compares corresponding elements ina
andb
to see if either isNaN
. - _mm_cmpunord_ps⚠(x86 or x86-64) and
sse
Compares each of the four floats ina
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. - _mm_cmpunord_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the result of comparing both of the lower elements ofa
andb
toNaN
. If either is equal toNaN
then0xFFFFFFFFFFFFFFFF
is used and0
otherwise. - _mm_cmpunord_ss⚠(x86 or x86-64) and
sse
Checks if the lowestf32
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
. - _mm_comieq_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for equality. - _mm_comieq_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if they are equal, or0
otherwise. - _mm_comige_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for greater-than-or-equal. - _mm_comige_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if the value froma
is greater than or equal to the one fromb
, or0
otherwise. - _mm_comigt_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for greater-than. - _mm_comigt_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if the value froma
is greater than the one fromb
, or0
otherwise. - _mm_comile_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for less-than-or-equal. - _mm_comile_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if the value froma
is less than or equal to the one fromb
, or0
otherwise. - _mm_comilt_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for less-than. - _mm_comilt_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if the value froma
is less than the one fromb
, or0
otherwise. - _mm_comineq_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for not-equal. - _mm_comineq_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if they are not equal, or0
otherwise. - _mm_crc32_u8⚠(x86 or x86-64) and
sse4.2
Starting with the initial value incrc
, return the accumulated CRC32-C value for unsigned 8-bit integerv
. - _mm_crc32_u16⚠(x86 or x86-64) and
sse4.2
Starting with the initial value incrc
, return the accumulated CRC32-C value for unsigned 16-bit integerv
. - _mm_crc32_u32⚠(x86 or x86-64) and
sse4.2
Starting with the initial value incrc
, return the accumulated CRC32-C value for unsigned 32-bit integerv
. - _mm_cvt_si2ss⚠(x86 or x86-64) and
sse
Alias for_mm_cvtsi32_ss
. - _mm_cvt_ss2si⚠(x86 or x86-64) and
sse
Alias for_mm_cvtss_si32
. - _mm_cvtepi8_epi16⚠(x86 or x86-64) and
sse4.1
Sign extend packed 8-bit integers ina
to packed 16-bit integers - _mm_cvtepi8_epi32⚠(x86 or x86-64) and
sse4.1
Sign extend packed 8-bit integers ina
to packed 32-bit integers - _mm_cvtepi8_epi64⚠(x86 or x86-64) and
sse4.1
Sign extend packed 8-bit integers in the low 8 bytes ofa
to packed 64-bit integers - _mm_cvtepi16_epi32⚠(x86 or x86-64) and
sse4.1
Sign extend packed 16-bit integers ina
to packed 32-bit integers - _mm_cvtepi16_epi64⚠(x86 or x86-64) and
sse4.1
Sign extend packed 16-bit integers ina
to packed 64-bit integers - _mm_cvtepi32_epi64⚠(x86 or x86-64) and
sse4.1
Sign extend packed 32-bit integers ina
to packed 64-bit integers - _mm_cvtepi32_pd⚠(x86 or x86-64) and
sse2
Converts the lower two packed 32-bit integers ina
to packed double-precision (64-bit) floating-point elements. - _mm_cvtepi32_ps⚠(x86 or x86-64) and
sse2
Converts packed 32-bit integers ina
to packed single-precision (32-bit) floating-point elements. - _mm_cvtepu8_epi16⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 8-bit integers ina
to packed 16-bit integers - _mm_cvtepu8_epi32⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 8-bit integers ina
to packed 32-bit integers - _mm_cvtepu8_epi64⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 8-bit integers ina
to packed 64-bit integers - _mm_cvtepu16_epi32⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 16-bit integers ina
to packed 32-bit integers - _mm_cvtepu16_epi64⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 16-bit integers ina
to packed 64-bit integers - _mm_cvtepu32_epi64⚠(x86 or x86-64) and
sse4.1
Zeroes extend packed unsigned 32-bit integers ina
to packed 64-bit integers - _mm_cvtpd_epi32⚠(x86 or x86-64) and
sse2
Converts packed double-precision (64-bit) floating-point elements ina
to packed 32-bit integers. - _mm_cvtpd_ps⚠(x86 or x86-64) and
sse2
Converts packed double-precision (64-bit) floating-point elements ina
to packed single-precision (32-bit) floating-point elements - _mm_cvtph_ps⚠(x86 or x86-64) and
f16c
Converts the 4 x 16-bit half-precision float values in the lowest 64-bit of the 128-bit vectora
into 4 x 32-bit float values stored in a 128-bit wide vector. - _mm_cvtps_epi32⚠(x86 or x86-64) and
sse2
Converts packed single-precision (32-bit) floating-point elements ina
to packed 32-bit integers. - _mm_cvtps_pd⚠(x86 or x86-64) and
sse2
Converts packed single-precision (32-bit) floating-point elements ina
to packed double-precision (64-bit) floating-point elements. - _mm_cvtps_ph⚠(x86 or x86-64) and
f16c
Converts the 4 x 32-bit float values in the 128-bit vectora
into 4 x 16-bit half-precision float values stored in the lowest 64-bit of a 128-bit vector. - _mm_cvtsd_f64⚠(x86 or x86-64) and
sse2
Returns the lower double-precision (64-bit) floating-point element ofa
. - _mm_cvtsd_si32⚠(x86 or x86-64) and
sse2
Converts the lower double-precision (64-bit) floating-point element in a to a 32-bit integer. - _mm_cvtsd_ss⚠(x86 or x86-64) and
sse2
Converts the lower double-precision (64-bit) floating-point element inb
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. - _mm_cvtsi32_sd⚠(x86 or x86-64) and
sse2
Returnsa
with its lower element replaced byb
after converting it to anf64
. - _mm_cvtsi32_si128⚠(x86 or x86-64) and
sse2
Returns a vector whose lowest element isa
and all higher elements are0
. - _mm_cvtsi32_ss⚠(x86 or x86-64) and
sse
Converts a 32 bit integer to a 32 bit float. The result vector is the input vectora
with the lowest 32 bit float replaced by the converted integer. - _mm_cvtsi128_si32⚠(x86 or x86-64) and
sse2
Returns the lowest element ofa
. - _mm_cvtss_f32⚠(x86 or x86-64) and
sse
Extracts the lowest 32 bit float from the input vector. - _mm_cvtss_sd⚠(x86 or x86-64) and
sse2
Converts the lower single-precision (32-bit) floating-point element inb
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. - _mm_cvtss_si32⚠(x86 or x86-64) and
sse
Converts the lowest 32 bit float in the input vector to a 32 bit integer. - _mm_cvtt_ss2si⚠(x86 or x86-64) and
sse
Alias for_mm_cvttss_si32
. - _mm_cvttpd_epi32⚠(x86 or x86-64) and
sse2
Converts packed double-precision (64-bit) floating-point elements ina
to packed 32-bit integers with truncation. - _mm_cvttps_epi32⚠(x86 or x86-64) and
sse2
Converts packed single-precision (32-bit) floating-point elements ina
to packed 32-bit integers with truncation. - _mm_cvttsd_si32⚠(x86 or x86-64) and
sse2
Converts the lower double-precision (64-bit) floating-point element ina
to a 32-bit integer with truncation. - _mm_cvttss_si32⚠(x86 or x86-64) and
sse
Converts the lowest 32 bit float in the input vector to a 32 bit integer with truncation. - _mm_div_pd⚠(x86 or x86-64) and
sse2
Divide packed double-precision (64-bit) floating-point elements ina
by packed elements inb
. - _mm_div_ps⚠(x86 or x86-64) and
sse
Divides __m128 vectors. - _mm_div_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the result of diving the lower element ofa
by the lower element ofb
. - _mm_div_ss⚠(x86 or x86-64) and
sse
Divides the first component ofb
bya
, the other components are copied froma
. - _mm_dp_pd⚠(x86 or x86-64) and
sse4.1
Returns the dot product of two __m128d vectors. - _mm_dp_ps⚠(x86 or x86-64) and
sse4.1
Returns the dot product of two __m128 vectors. - _mm_extract_epi8⚠(x86 or x86-64) and
sse4.1
Extracts an 8-bit integer froma
, selected withIMM8
. Returns a 32-bit integer containing the zero-extended integer data. - _mm_extract_epi16⚠(x86 or x86-64) and
sse2
Returns theimm8
element ofa
. - _mm_extract_epi32⚠(x86 or x86-64) and
sse4.1
Extracts an 32-bit integer froma
selected withIMM8
- _mm_extract_ps⚠(x86 or x86-64) and
sse4.1
Extracts a single-precision (32-bit) floating-point element froma
, selected withIMM8
. The returnedi32
stores the float’s bit-pattern, and may be converted back to a floating point number via casting. - _mm_extract_si64⚠(x86 or x86-64) and
sse4a
Extracts the bit range specified byy
from the lower 64 bits ofx
. - _mm_floor_pd⚠(x86 or x86-64) and
sse4.1
Round the packed double-precision (64-bit) floating-point elements ina
down to an integer value, and stores the results as packed double-precision floating-point elements. - _mm_floor_ps⚠(x86 or x86-64) and
sse4.1
Round the packed single-precision (32-bit) floating-point elements ina
down to an integer value, and stores the results as packed single-precision floating-point elements. - _mm_floor_sd⚠(x86 or x86-64) and
sse4.1
Round the lower double-precision (64-bit) floating-point element inb
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. - _mm_floor_ss⚠(x86 or x86-64) and
sse4.1
Round the lower single-precision (32-bit) floating-point element inb
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. - _mm_fmadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and add the intermediate result to packed elements inc
. - _mm_fmadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and add the intermediate result to packed elements inc
. - _mm_fmadd_sd⚠(x86 or x86-64) and
fma
Multiplies the lower double-precision (64-bit) floating-point elements ina
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. - _mm_fmadd_ss⚠(x86 or x86-64) and
fma
Multiplies the lower single-precision (32-bit) floating-point elements ina
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. - _mm_fmaddsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - _mm_fmaddsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and alternatively add and subtract packed elements inc
to/from the intermediate result. - _mm_fmsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the intermediate result. - _mm_fmsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the intermediate result. - _mm_fmsub_sd⚠(x86 or x86-64) and
fma
Multiplies the lower double-precision (64-bit) floating-point elements ina
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. - _mm_fmsub_ss⚠(x86 or x86-64) and
fma
Multiplies the lower single-precision (32-bit) floating-point elements ina
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. - _mm_fmsubadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - _mm_fmsubadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and alternatively subtract and add packed elements inc
from/to the intermediate result. - _mm_fnmadd_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and add the negated intermediate result to packed elements inc
. - _mm_fnmadd_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and add the negated intermediate result to packed elements inc
. - _mm_fnmadd_sd⚠(x86 or x86-64) and
fma
Multiplies the lower double-precision (64-bit) floating-point elements ina
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. - _mm_fnmadd_ss⚠(x86 or x86-64) and
fma
Multiplies the lower single-precision (32-bit) floating-point elements ina
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. - _mm_fnmsub_pd⚠(x86 or x86-64) and
fma
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the negated intermediate result. - _mm_fnmsub_ps⚠(x86 or x86-64) and
fma
Multiplies packed single-precision (32-bit) floating-point elements ina
andb
, and subtract packed elements inc
from the negated intermediate result. - _mm_fnmsub_sd⚠(x86 or x86-64) and
fma
Multiplies the lower double-precision (64-bit) floating-point elements ina
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. - _mm_fnmsub_ss⚠(x86 or x86-64) and
fma
Multiplies the lower single-precision (32-bit) floating-point elements ina
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. - Gets the unsigned 32-bit value of the MXCSR control and status register.
- _mm_hadd_epi16⚠(x86 or x86-64) and
ssse3
Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of[8 x i16]
. - _mm_hadd_epi32⚠(x86 or x86-64) and
ssse3
Horizontally adds the adjacent pairs of values contained in 2 packed 128-bit vectors of[4 x i32]
. - _mm_hadd_pd⚠(x86 or x86-64) and
sse3
Horizontally adds adjacent pairs of double-precision (64-bit) floating-point elements ina
andb
, and pack the results. - _mm_hadd_ps⚠(x86 or x86-64) and
sse3
Horizontally adds adjacent pairs of single-precision (32-bit) floating-point elements ina
andb
, and pack the results. - _mm_hadds_epi16⚠(x86 or x86-64) and
ssse3
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. - _mm_hsub_epi16⚠(x86 or x86-64) and
ssse3
Horizontally subtract the adjacent pairs of values contained in 2 packed 128-bit vectors of[8 x i16]
. - _mm_hsub_epi32⚠(x86 or x86-64) and
ssse3
Horizontally subtract the adjacent pairs of values contained in 2 packed 128-bit vectors of[4 x i32]
. - _mm_hsub_pd⚠(x86 or x86-64) and
sse3
Horizontally subtract adjacent pairs of double-precision (64-bit) floating-point elements ina
andb
, and pack the results. - _mm_hsub_ps⚠(x86 or x86-64) and
sse3
Horizontally adds adjacent pairs of single-precision (32-bit) floating-point elements ina
andb
, and pack the results. - _mm_hsubs_epi16⚠(x86 or x86-64) and
ssse3
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. - _mm_i32gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i32gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i32gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i32gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i64gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i64gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i64gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_i64gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
at offsets determined byoffsets * scale
, wherescale
should be 1, 2, 4 or 8. - _mm_insert_epi8⚠(x86 or x86-64) and
sse4.1
Returns a copy ofa
with the 8-bit integer fromi
inserted at a location specified byIMM8
. - _mm_insert_epi16⚠(x86 or x86-64) and
sse2
Returns a new vector where theimm8
element ofa
is replaced withi
. - _mm_insert_epi32⚠(x86 or x86-64) and
sse4.1
Returns a copy ofa
with the 32-bit integer fromi
inserted at a location specified byIMM8
. - _mm_insert_ps⚠(x86 or x86-64) and
sse4.1
Select a single value inb
to store at some position ina
, Then zero elements according toIMM8
. - _mm_insert_si64⚠(x86 or x86-64) and
sse4a
Inserts the[length:0]
bits ofy
intox
atindex
. - _mm_lddqu_si128⚠(x86 or x86-64) and
sse3
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. - _mm_lfence⚠(x86 or x86-64) and
sse2
Performs a serializing operation on all load-from-memory instructions that were issued prior to this instruction. - _mm_load1_pd⚠(x86 or x86-64) and
sse2
Loads a double-precision (64-bit) floating-point element from memory into both elements of returned vector. - _mm_load1_ps⚠(x86 or x86-64) and
sse
Construct a__m128
by duplicating the value read fromp
into all elements. - _mm_load_pd⚠(x86 or x86-64) and
sse2
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. - _mm_load_pd1⚠(x86 or x86-64) and
sse2
Loads a double-precision (64-bit) floating-point element from memory into both elements of returned vector. - _mm_load_ps⚠(x86 or x86-64) and
sse
Loads fourf32
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). - _mm_load_ps1⚠(x86 or x86-64) and
sse
Alias for_mm_load1_ps
- _mm_load_sd⚠(x86 or x86-64) and
sse2
Loads a 64-bit double-precision value to the low element of a 128-bit integer vector and clears the upper element. - _mm_load_si128⚠(x86 or x86-64) and
sse2
Loads 128-bits of integer data from memory into a new vector. - _mm_load_ss⚠(x86 or x86-64) and
sse
Construct a__m128
with the lowest element read fromp
and the other elements set to zero. - _mm_loaddup_pd⚠(x86 or x86-64) and
sse3
Loads a double-precision (64-bit) floating-point element from memory into both elements of return vector. - _mm_loadh_pd⚠(x86 or x86-64) and
sse2
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. - _mm_loadl_epi64⚠(x86 or x86-64) and
sse2
Loads 64-bit integer from memory into first element of returned vector. - _mm_loadl_pd⚠(x86 or x86-64) and
sse2
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. - _mm_loadr_pd⚠(x86 or x86-64) and
sse2
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. - _mm_loadr_ps⚠(x86 or x86-64) and
sse
Loads fourf32
values from aligned memory into a__m128
in reverse order. - _mm_loadu_pd⚠(x86 or x86-64) and
sse2
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. - _mm_loadu_ps⚠(x86 or x86-64) and
sse
Loads fourf32
values from memory into a__m128
. There are no restrictions on memory alignment. For aligned memory_mm_load_ps
may be faster. - _mm_loadu_si64⚠(x86 or x86-64) and
sse
Loads unaligned 64-bits of integer data from memory into new vector. - _mm_loadu_si128⚠(x86 or x86-64) and
sse2
Loads 128-bits of integer data from memory into a new vector. - _mm_madd_epi16⚠(x86 or x86-64) and
sse2
Multiplies and then horizontally add signed 16 bit integers ina
andb
. - _mm_maddubs_epi16⚠(x86 or x86-64) and
ssse3
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. - _mm_mask_i32gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i32gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i32gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i32gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i64gather_epi32⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i64gather_epi64⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i64gather_pd⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_mask_i64gather_ps⚠(x86 or x86-64) and
avx2
Returns values fromslice
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. - _mm_maskload_epi32⚠(x86 or x86-64) and
avx2
Loads packed 32-bit integers from memory pointed bymem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - _mm_maskload_epi64⚠(x86 or x86-64) and
avx2
Loads packed 64-bit integers from memory pointed bymem_addr
usingmask
(elements are zeroed out when the highest bit is not set in the corresponding element). - _mm_maskload_pd⚠(x86 or x86-64) and
avx
Loads packed double-precision (64-bit) floating-point elements from memory into result usingmask
(elements are zeroed out when the high bit of the corresponding element is not set). - _mm_maskload_ps⚠(x86 or x86-64) and
avx
Loads packed single-precision (32-bit) floating-point elements from memory into result usingmask
(elements are zeroed out when the high bit of the corresponding element is not set). - _mm_maskmoveu_si128⚠(x86 or x86-64) and
sse2
Conditionally store 8-bit integer elements froma
into memory usingmask
. - _mm_maskstore_epi32⚠(x86 or x86-64) and
avx2
Stores packed 32-bit integers froma
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - _mm_maskstore_epi64⚠(x86 or x86-64) and
avx2
Stores packed 64-bit integers froma
into memory pointed bymem_addr
usingmask
(elements are not stored when the highest bit is not set in the corresponding element). - _mm_maskstore_pd⚠(x86 or x86-64) and
avx
Stores packed double-precision (64-bit) floating-point elements froma
into memory usingmask
. - _mm_maskstore_ps⚠(x86 or x86-64) and
avx
Stores packed single-precision (32-bit) floating-point elements froma
into memory usingmask
. - _mm_max_epi8⚠(x86 or x86-64) and
sse4.1
Compares packed 8-bit integers ina
andb
and returns packed maximum values in dst. - _mm_max_epi16⚠(x86 or x86-64) and
sse2
Compares packed 16-bit integers ina
andb
, and returns the packed maximum values. - _mm_max_epi32⚠(x86 or x86-64) and
sse4.1
Compares packed 32-bit integers ina
andb
, and returns packed maximum values. - _mm_max_epu8⚠(x86 or x86-64) and
sse2
Compares packed unsigned 8-bit integers ina
andb
, and returns the packed maximum values. - _mm_max_epu16⚠(x86 or x86-64) and
sse4.1
Compares packed unsigned 16-bit integers ina
andb
, and returns packed maximum. - _mm_max_epu32⚠(x86 or x86-64) and
sse4.1
Compares packed unsigned 32-bit integers ina
andb
, and returns packed maximum values. - _mm_max_pd⚠(x86 or x86-64) and
sse2
Returns a new vector with the maximum values from corresponding elements ina
andb
. - _mm_max_ps⚠(x86 or x86-64) and
sse
Compares packed single-precision (32-bit) floating-point elements ina
andb
, and return the corresponding maximum values. - _mm_max_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the maximum of the lower elements ofa
andb
. - _mm_max_ss⚠(x86 or x86-64) and
sse
Compares the first single-precision (32-bit) floating-point element ofa
andb
, and return the maximum value in the first element of the return value, the other elements are copied froma
. - _mm_mfence⚠(x86 or x86-64) and
sse2
Performs a serializing operation on all load-from-memory and store-to-memory instructions that were issued prior to this instruction. - _mm_min_epi8⚠(x86 or x86-64) and
sse4.1
Compares packed 8-bit integers ina
andb
and returns packed minimum values in dst. - _mm_min_epi16⚠(x86 or x86-64) and
sse2
Compares packed 16-bit integers ina
andb
, and returns the packed minimum values. - _mm_min_epi32⚠(x86 or x86-64) and
sse4.1
Compares packed 32-bit integers ina
andb
, and returns packed minimum values. - _mm_min_epu8⚠(x86 or x86-64) and
sse2
Compares packed unsigned 8-bit integers ina
andb
, and returns the packed minimum values. - _mm_min_epu16⚠(x86 or x86-64) and
sse4.1
Compares packed unsigned 16-bit integers ina
andb
, and returns packed minimum. - _mm_min_epu32⚠(x86 or x86-64) and
sse4.1
Compares packed unsigned 32-bit integers ina
andb
, and returns packed minimum values. - _mm_min_pd⚠(x86 or x86-64) and
sse2
Returns a new vector with the minimum values from corresponding elements ina
andb
. - _mm_min_ps⚠(x86 or x86-64) and
sse
Compares packed single-precision (32-bit) floating-point elements ina
andb
, and return the corresponding minimum values. - _mm_min_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the minimum of the lower elements ofa
andb
. - _mm_min_ss⚠(x86 or x86-64) and
sse
Compares the first single-precision (32-bit) floating-point element ofa
andb
, and return the minimum value in the first element of the return value, the other elements are copied froma
. - _mm_minpos_epu16⚠(x86 or x86-64) and
sse4.1
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. - _mm_move_epi64⚠(x86 or x86-64) and
sse2
Returns a vector where the low element is extracted froma
and its upper element is zero. - _mm_move_sd⚠(x86 or x86-64) and
sse2
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. - _mm_move_ss⚠(x86 or x86-64) and
sse
Returns a__m128
with the first component fromb
and the remaining components froma
. - _mm_movedup_pd⚠(x86 or x86-64) and
sse3
Duplicate the low double-precision (64-bit) floating-point element froma
. - _mm_movehdup_ps⚠(x86 or x86-64) and
sse3
Duplicate odd-indexed single-precision (32-bit) floating-point elements froma
. - _mm_movehl_ps⚠(x86 or x86-64) and
sse
Combine higher half ofa
andb
. The higher half ofb
occupies the lower half of result. - _mm_moveldup_ps⚠(x86 or x86-64) and
sse3
Duplicate even-indexed single-precision (32-bit) floating-point elements froma
. - _mm_movelh_ps⚠(x86 or x86-64) and
sse
Combine lower half ofa
andb
. The lower half ofb
occupies the higher half of result. - _mm_movemask_epi8⚠(x86 or x86-64) and
sse2
Returns a mask of the most significant bit of each element ina
. - _mm_movemask_pd⚠(x86 or x86-64) and
sse2
Returns a mask of the most significant bit of each element ina
. - _mm_movemask_ps⚠(x86 or x86-64) and
sse
Returns a mask of the most significant bit of each element ina
. - _mm_mpsadbw_epu8⚠(x86 or x86-64) and
sse4.1
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. - _mm_mul_epi32⚠(x86 or x86-64) and
sse4.1
Multiplies the low 32-bit integers from each packed 64-bit element ina
andb
, and returns the signed 64-bit result. - _mm_mul_epu32⚠(x86 or x86-64) and
sse2
Multiplies the low unsigned 32-bit integers from each packed 64-bit element ina
andb
. - _mm_mul_pd⚠(x86 or x86-64) and
sse2
Multiplies packed double-precision (64-bit) floating-point elements ina
andb
. - _mm_mul_ps⚠(x86 or x86-64) and
sse
Multiplies __m128 vectors. - _mm_mul_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by multiplying the low elements ofa
andb
. - _mm_mul_ss⚠(x86 or x86-64) and
sse
Multiplies the first component ofa
andb
, the other components are copied froma
. - _mm_mulhi_epi16⚠(x86 or x86-64) and
sse2
Multiplies the packed 16-bit integers ina
andb
. - _mm_mulhi_epu16⚠(x86 or x86-64) and
sse2
Multiplies the packed unsigned 16-bit integers ina
andb
. - _mm_mulhrs_epi16⚠(x86 or x86-64) and
ssse3
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. - _mm_mullo_epi16⚠(x86 or x86-64) and
sse2
Multiplies the packed 16-bit integers ina
andb
. - _mm_mullo_epi32⚠(x86 or x86-64) and
sse4.1
Multiplies the packed 32-bit integers ina
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. - _mm_or_pd⚠(x86 or x86-64) and
sse2
Computes the bitwise OR ofa
andb
. - _mm_or_ps⚠(x86 or x86-64) and
sse
Bitwise OR of packed single-precision (32-bit) floating-point elements. - _mm_or_si128⚠(x86 or x86-64) and
sse2
Computes the bitwise OR of 128 bits (representing integer data) ina
andb
. - _mm_packs_epi16⚠(x86 or x86-64) and
sse2
Converts packed 16-bit integers froma
andb
to packed 8-bit integers using signed saturation. - _mm_packs_epi32⚠(x86 or x86-64) and
sse2
Converts packed 32-bit integers froma
andb
to packed 16-bit integers using signed saturation. - _mm_packus_epi16⚠(x86 or x86-64) and
sse2
Converts packed 16-bit integers froma
andb
to packed 8-bit integers using unsigned saturation. - _mm_packus_epi32⚠(x86 or x86-64) and
sse4.1
Converts packed 32-bit integers froma
andb
to packed 16-bit integers using unsigned saturation - _mm_pause⚠x86 or x86-64Provides a hint to the processor that the code sequence is a spin-wait loop.
- _mm_permute_pd⚠(x86 or x86-64) and
avx,sse2
Shuffles double-precision (64-bit) floating-point elements ina
using the control inimm8
. - _mm_permute_ps⚠(x86 or x86-64) and
avx,sse
Shuffles single-precision (32-bit) floating-point elements ina
using the control inimm8
. - _mm_permutevar_pd⚠(x86 or x86-64) and
avx
Shuffles double-precision (64-bit) floating-point elements ina
using the control inb
. - _mm_permutevar_ps⚠(x86 or x86-64) and
avx
Shuffles single-precision (32-bit) floating-point elements ina
using the control inb
. - _mm_prefetch⚠(x86 or x86-64) and
sse
Fetch the cache line that contains addressp
using the givenSTRATEGY
. - _mm_rcp_ps⚠(x86 or x86-64) and
sse
Returns the approximate reciprocal of packed single-precision (32-bit) floating-point elements ina
. - _mm_rcp_ss⚠(x86 or x86-64) and
sse
Returns the approximate reciprocal of the first single-precision (32-bit) floating-point element ina
, the other elements are unchanged. - _mm_round_pd⚠(x86 or x86-64) and
sse4.1
Round the packed double-precision (64-bit) floating-point elements ina
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: - _mm_round_ps⚠(x86 or x86-64) and
sse4.1
Round the packed single-precision (32-bit) floating-point elements ina
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: - _mm_round_sd⚠(x86 or x86-64) and
sse4.1
Round the lower double-precision (64-bit) floating-point element inb
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: - _mm_round_ss⚠(x86 or x86-64) and
sse4.1
Round the lower single-precision (32-bit) floating-point element inb
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: - _mm_rsqrt_ps⚠(x86 or x86-64) and
sse
Returns the approximate reciprocal square root of packed single-precision (32-bit) floating-point elements ina
. - _mm_rsqrt_ss⚠(x86 or x86-64) and
sse
Returns the approximate reciprocal square root of the first single-precision (32-bit) floating-point element ina
, the other elements are unchanged. - _mm_sad_epu8⚠(x86 or x86-64) and
sse2
Sum the absolute differences of packed unsigned 8-bit integers. - _mm_set1_epi8⚠(x86 or x86-64) and
sse2
Broadcasts 8-bit integera
to all elements. - _mm_set1_epi16⚠(x86 or x86-64) and
sse2
Broadcasts 16-bit integera
to all elements. - _mm_set1_epi32⚠(x86 or x86-64) and
sse2
Broadcasts 32-bit integera
to all elements. - _mm_set1_epi64x⚠(x86 or x86-64) and
sse2
Broadcasts 64-bit integera
to all elements. - _mm_set1_pd⚠(x86 or x86-64) and
sse2
Broadcasts double-precision (64-bit) floating-point value a to all elements of the return value. - _mm_set1_ps⚠(x86 or x86-64) and
sse
Construct a__m128
with all element set toa
. - _mm_set_epi8⚠(x86 or x86-64) and
sse2
Sets packed 8-bit integers with the supplied values. - _mm_set_epi16⚠(x86 or x86-64) and
sse2
Sets packed 16-bit integers with the supplied values. - _mm_set_epi32⚠(x86 or x86-64) and
sse2
Sets packed 32-bit integers with the supplied values. - _mm_set_epi64x⚠(x86 or x86-64) and
sse2
Sets packed 64-bit integers with the supplied values, from highest to lowest. - _mm_set_pd⚠(x86 or x86-64) and
sse2
Sets packed double-precision (64-bit) floating-point elements in the return value with the supplied values. - _mm_set_pd1⚠(x86 or x86-64) and
sse2
Broadcasts double-precision (64-bit) floating-point value a to all elements of the return value. - _mm_set_ps⚠(x86 or x86-64) and
sse
Construct a__m128
from four floating point values highest to lowest. - _mm_set_ps1⚠(x86 or x86-64) and
sse
Alias for_mm_set1_ps
- _mm_set_sd⚠(x86 or x86-64) and
sse2
Copies double-precision (64-bit) floating-point elementa
to the lower element of the packed 64-bit return value. - _mm_set_ss⚠(x86 or x86-64) and
sse
Construct a__m128
with the lowest element set toa
and the rest set to zero. - Sets the MXCSR register with the 32-bit unsigned integer value.
- _mm_setr_epi8⚠(x86 or x86-64) and
sse2
Sets packed 8-bit integers with the supplied values in reverse order. - _mm_setr_epi16⚠(x86 or x86-64) and
sse2
Sets packed 16-bit integers with the supplied values in reverse order. - _mm_setr_epi32⚠(x86 or x86-64) and
sse2
Sets packed 32-bit integers with the supplied values in reverse order. - _mm_setr_pd⚠(x86 or x86-64) and
sse2
Sets packed double-precision (64-bit) floating-point elements in the return value with the supplied values in reverse order. - _mm_setr_ps⚠(x86 or x86-64) and
sse
Construct a__m128
from four floating point values lowest to highest. - _mm_setzero_pd⚠(x86 or x86-64) and
sse2
Returns packed double-precision (64-bit) floating-point elements with all zeros. - _mm_setzero_ps⚠(x86 or x86-64) and
sse
Construct a__m128
with all elements initialized to zero. - _mm_setzero_si128⚠(x86 or x86-64) and
sse2
Returns a vector with all elements set to zero. - _mm_sfence⚠(x86 or x86-64) and
sse
Performs a serializing operation on all non-temporal (“streaming”) store instructions that were issued by the current thread prior to this instruction. - _mm_sha1msg1_epu32⚠(x86 or x86-64) and
sha
Performs an intermediate calculation for the next four SHA1 message values (unsigned 32-bit integers) using previous message values froma
andb
, and returning the result. - _mm_sha1msg2_epu32⚠(x86 or x86-64) and
sha
Performs the final calculation for the next four SHA1 message values (unsigned 32-bit integers) using the intermediate result ina
and the previous message values inb
, and returns the result. - _mm_sha1nexte_epu32⚠(x86 or x86-64) and
sha
Calculate SHA1 state variable E after four rounds of operation from the current SHA1 state variablea
, add that value to the scheduled values (unsigned 32-bit integers) inb
, and returns the result. - _mm_sha1rnds4_epu32⚠(x86 or x86-64) and
sha
Performs four rounds of SHA1 operation using an initial SHA1 state (A,B,C,D) froma
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. - _mm_sha256msg1_epu32⚠(x86 or x86-64) and
sha
Performs an intermediate calculation for the next four SHA256 message values (unsigned 32-bit integers) using previous message values froma
andb
, and return the result. - _mm_sha256msg2_epu32⚠(x86 or x86-64) and
sha
Performs the final calculation for the next four SHA256 message values (unsigned 32-bit integers) using previous message values froma
andb
, and return the result. - _mm_sha256rnds2_epu32⚠(x86 or x86-64) and
sha
Performs 2 rounds of SHA256 operation using an initial SHA256 state (C,D,G,H) froma
, 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. - _mm_shuffle_epi8⚠(x86 or x86-64) and
ssse3
Shuffles bytes froma
according to the content ofb
. - _mm_shuffle_epi32⚠(x86 or x86-64) and
sse2
Shuffles 32-bit integers ina
using the control inIMM8
. - _mm_shuffle_pd⚠(x86 or x86-64) and
sse2
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. - _mm_shuffle_ps⚠(x86 or x86-64) and
sse
Shuffles packed single-precision (32-bit) floating-point elements ina
andb
usingMASK
. - _mm_shufflehi_epi16⚠(x86 or x86-64) and
sse2
Shuffles 16-bit integers in the high 64 bits ofa
using the control inIMM8
. - _mm_shufflelo_epi16⚠(x86 or x86-64) and
sse2
Shuffles 16-bit integers in the low 64 bits ofa
using the control inIMM8
. - _mm_sign_epi8⚠(x86 or x86-64) and
ssse3
Negates packed 8-bit integers ina
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. - _mm_sign_epi16⚠(x86 or x86-64) and
ssse3
Negates packed 16-bit integers ina
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. - _mm_sign_epi32⚠(x86 or x86-64) and
ssse3
Negates packed 32-bit integers ina
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. - _mm_sll_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
left bycount
while shifting in zeros. - _mm_sll_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
left bycount
while shifting in zeros. - _mm_sll_epi64⚠(x86 or x86-64) and
sse2
Shifts packed 64-bit integers ina
left bycount
while shifting in zeros. - _mm_slli_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
left byIMM8
while shifting in zeros. - _mm_slli_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
left byIMM8
while shifting in zeros. - _mm_slli_epi64⚠(x86 or x86-64) and
sse2
Shifts packed 64-bit integers ina
left byIMM8
while shifting in zeros. - _mm_slli_si128⚠(x86 or x86-64) and
sse2
Shiftsa
left byIMM8
bytes while shifting in zeros. - _mm_sllv_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - _mm_sllv_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
left by the amount specified by the corresponding element incount
while shifting in zeros, and returns the result. - _mm_sqrt_pd⚠(x86 or x86-64) and
sse2
Returns a new vector with the square root of each of the values ina
. - _mm_sqrt_ps⚠(x86 or x86-64) and
sse
Returns the square root of packed single-precision (32-bit) floating-point elements ina
. - _mm_sqrt_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by the square root of the lower elementb
. - _mm_sqrt_ss⚠(x86 or x86-64) and
sse
Returns the square root of the first single-precision (32-bit) floating-point element ina
, the other elements are unchanged. - _mm_sra_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
right bycount
while shifting in sign bits. - _mm_sra_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
right bycount
while shifting in sign bits. - _mm_srai_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
right byIMM8
while shifting in sign bits. - _mm_srai_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
right byIMM8
while shifting in sign bits. - _mm_srav_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in sign bits. - _mm_srl_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
right bycount
while shifting in zeros. - _mm_srl_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
right bycount
while shifting in zeros. - _mm_srl_epi64⚠(x86 or x86-64) and
sse2
Shifts packed 64-bit integers ina
right bycount
while shifting in zeros. - _mm_srli_epi16⚠(x86 or x86-64) and
sse2
Shifts packed 16-bit integers ina
right byIMM8
while shifting in zeros. - _mm_srli_epi32⚠(x86 or x86-64) and
sse2
Shifts packed 32-bit integers ina
right byIMM8
while shifting in zeros. - _mm_srli_epi64⚠(x86 or x86-64) and
sse2
Shifts packed 64-bit integers ina
right byIMM8
while shifting in zeros. - _mm_srli_si128⚠(x86 or x86-64) and
sse2
Shiftsa
right byIMM8
bytes while shifting in zeros. - _mm_srlv_epi32⚠(x86 or x86-64) and
avx2
Shifts packed 32-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in zeros, - _mm_srlv_epi64⚠(x86 or x86-64) and
avx2
Shifts packed 64-bit integers ina
right by the amount specified by the corresponding element incount
while shifting in zeros, - _mm_store1_pd⚠(x86 or x86-64) and
sse2
Stores the lower double-precision (64-bit) floating-point element froma
into 2 contiguous elements in memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_store1_ps⚠(x86 or x86-64) and
sse
Stores the lowest 32 bit float ofa
repeated four times into aligned memory. - _mm_store_pd⚠(x86 or x86-64) and
sse2
Stores 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) froma
into memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_store_pd1⚠(x86 or x86-64) and
sse2
Stores the lower double-precision (64-bit) floating-point element froma
into 2 contiguous elements in memory.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_store_ps⚠(x86 or x86-64) and
sse
Stores four 32-bit floats into aligned memory. - _mm_store_ps1⚠(x86 or x86-64) and
sse
Alias for_mm_store1_ps
- _mm_store_sd⚠(x86 or x86-64) and
sse2
Stores the lower 64 bits of a 128-bit vector of[2 x double]
to a memory location. - _mm_store_si128⚠(x86 or x86-64) and
sse2
Stores 128-bits of integer data froma
into memory. - _mm_store_ss⚠(x86 or x86-64) and
sse
Stores the lowest 32 bit float ofa
into memory. - _mm_storeh_pd⚠(x86 or x86-64) and
sse2
Stores the upper 64 bits of a 128-bit vector of[2 x double]
to a memory location. - _mm_storel_epi64⚠(x86 or x86-64) and
sse2
Stores the lower 64-bit integera
to a memory location. - _mm_storel_pd⚠(x86 or x86-64) and
sse2
Stores the lower 64 bits of a 128-bit vector of[2 x double]
to a memory location. - _mm_storer_pd⚠(x86 or x86-64) and
sse2
Stores 2 double-precision (64-bit) floating-point elements froma
into memory in reverse order.mem_addr
must be aligned on a 16-byte boundary or a general-protection exception may be generated. - _mm_storer_ps⚠(x86 or x86-64) and
sse
Stores four 32-bit floats into aligned memory in reverse order. - _mm_storeu_pd⚠(x86 or x86-64) and
sse2
Stores 128-bits (composed of 2 packed double-precision (64-bit) floating-point elements) froma
into memory.mem_addr
does not need to be aligned on any particular boundary. - _mm_storeu_ps⚠(x86 or x86-64) and
sse
Stores four 32-bit floats into memory. There are no restrictions on memory alignment. For aligned memory_mm_store_ps
may be faster. - _mm_storeu_si128⚠(x86 or x86-64) and
sse2
Stores 128-bits of integer data froma
into memory. - _mm_stream_pd⚠(x86 or x86-64) and
sse2
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). - _mm_stream_ps⚠(x86 or x86-64) and
sse
Storesa
into the memory atmem_addr
using a non-temporal memory hint. - _mm_stream_sd⚠(x86 or x86-64) and
sse4a
Non-temporal store ofa.0
intop
. - _mm_stream_si32⚠(x86 or x86-64) and
sse2
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). - _mm_stream_si128⚠(x86 or x86-64) and
sse2
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). - _mm_stream_ss⚠(x86 or x86-64) and
sse4a
Non-temporal store ofa.0
intop
. - _mm_sub_epi8⚠(x86 or x86-64) and
sse2
Subtracts packed 8-bit integers inb
from packed 8-bit integers ina
. - _mm_sub_epi16⚠(x86 or x86-64) and
sse2
Subtracts packed 16-bit integers inb
from packed 16-bit integers ina
. - _mm_sub_epi32⚠(x86 or x86-64) and
sse2
Subtract packed 32-bit integers inb
from packed 32-bit integers ina
. - _mm_sub_epi64⚠(x86 or x86-64) and
sse2
Subtract packed 64-bit integers inb
from packed 64-bit integers ina
. - _mm_sub_pd⚠(x86 or x86-64) and
sse2
Subtract packed double-precision (64-bit) floating-point elements inb
froma
. - _mm_sub_ps⚠(x86 or x86-64) and
sse
Subtracts __m128 vectors. - _mm_sub_sd⚠(x86 or x86-64) and
sse2
Returns a new vector with the low element ofa
replaced by subtracting the low element byb
from the low element ofa
. - _mm_sub_ss⚠(x86 or x86-64) and
sse
Subtracts the first component ofb
froma
, the other components are copied froma
. - _mm_subs_epi8⚠(x86 or x86-64) and
sse2
Subtract packed 8-bit integers inb
from packed 8-bit integers ina
using saturation. - _mm_subs_epi16⚠(x86 or x86-64) and
sse2
Subtract packed 16-bit integers inb
from packed 16-bit integers ina
using saturation. - _mm_subs_epu8⚠(x86 or x86-64) and
sse2
Subtract packed unsigned 8-bit integers inb
from packed unsigned 8-bit integers ina
using saturation. - _mm_subs_epu16⚠(x86 or x86-64) and
sse2
Subtract packed unsigned 16-bit integers inb
from packed unsigned 16-bit integers ina
using saturation. - _mm_test_all_ones⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits ina
128-bit integer vector are all ones. - _mm_test_all_zeros⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits in a 128-bit integer vector are all zeros. - _mm_test_mix_ones_zeros⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits in a 128-bit integer vector are neither all zeros nor all ones. - _mm_testc_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm_testc_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm_testc_si128⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits in a 128-bit integer vector are all ones. - _mm_testnzc_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm_testnzc_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm_testnzc_si128⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits in a 128-bit integer vector are neither all zeros nor all ones. - _mm_testz_pd⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing double-precision (64-bit) floating-point elements) ina
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. - _mm_testz_ps⚠(x86 or x86-64) and
avx
Computes the bitwise AND of 128 bits (representing single-precision (32-bit) floating-point elements) ina
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. - _mm_testz_si128⚠(x86 or x86-64) and
sse4.1
Tests whether the specified bits in a 128-bit integer vector are all zeros. - _mm_tzcnt_32⚠(x86 or x86-64) and
bmi1
Counts the number of trailing least significant zero bits. - _mm_ucomieq_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for equality. - _mm_ucomieq_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if they are equal, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - _mm_ucomige_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for greater-than-or-equal. - _mm_ucomige_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
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. - _mm_ucomigt_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for greater-than. - _mm_ucomigt_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
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. - _mm_ucomile_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for less-than-or-equal. - _mm_ucomile_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
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. - _mm_ucomilt_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for less-than. - _mm_ucomilt_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
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. - _mm_ucomineq_sd⚠(x86 or x86-64) and
sse2
Compares the lower element ofa
andb
for not-equal. - _mm_ucomineq_ss⚠(x86 or x86-64) and
sse
Compares two 32-bit floats from the low-order bits ofa
andb
. Returns1
if they are not equal, or0
otherwise. This instruction will not signal an exception if either argument is a quiet NaN. - _mm_undefined_pd⚠(x86 or x86-64) and
sse2
Returns vector of type __m128d with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm_undefined_ps⚠(x86 or x86-64) and
sse
Returns vector of type __m128 with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm_undefined_si128⚠(x86 or x86-64) and
sse2
Returns vector of type __m128i with indeterminate elements. Despite being “undefined”, this is some valid value and not equivalent tomem::MaybeUninit
. In practice, this is equivalent tomem::zeroed
. - _mm_unpackhi_epi8⚠(x86 or x86-64) and
sse2
Unpacks and interleave 8-bit integers from the high half ofa
andb
. - _mm_unpackhi_epi16⚠(x86 or x86-64) and
sse2
Unpacks and interleave 16-bit integers from the high half ofa
andb
. - _mm_unpackhi_epi32⚠(x86 or x86-64) and
sse2
Unpacks and interleave 32-bit integers from the high half ofa
andb
. - _mm_unpackhi_epi64⚠(x86 or x86-64) and
sse2
Unpacks and interleave 64-bit integers from the high half ofa
andb
. - _mm_unpackhi_pd⚠(x86 or x86-64) and
sse2
The resulting__m128d
element is composed by the low-order values of the two__m128d
interleaved input elements, i.e.: - _mm_unpackhi_ps⚠(x86 or x86-64) and
sse
Unpacks and interleave single-precision (32-bit) floating-point elements from the higher half ofa
andb
. - _mm_unpacklo_epi8⚠(x86 or x86-64) and
sse2
Unpacks and interleave 8-bit integers from the low half ofa
andb
. - _mm_unpacklo_epi16⚠(x86 or x86-64) and
sse2
Unpacks and interleave 16-bit integers from the low half ofa
andb
. - _mm_unpacklo_epi32⚠(x86 or x86-64) and
sse2
Unpacks and interleave 32-bit integers from the low half ofa
andb
. - _mm_unpacklo_epi64⚠(x86 or x86-64) and
sse2
Unpacks and interleave 64-bit integers from the low half ofa
andb
. - _mm_unpacklo_pd⚠(x86 or x86-64) and
sse2
The resulting__m128d
element is composed by the high-order values of the two__m128d
interleaved input elements, i.e.: - _mm_unpacklo_ps⚠(x86 or x86-64) and
sse
Unpacks and interleave single-precision (32-bit) floating-point elements from the lower half ofa
andb
. - _mm_xor_pd⚠(x86 or x86-64) and
sse2
Computes the bitwise XOR ofa
andb
. - _mm_xor_ps⚠(x86 or x86-64) and
sse
Bitwise exclusive OR of packed single-precision (32-bit) floating-point elements. - _mm_xor_si128⚠(x86 or x86-64) and
sse2
Computes the bitwise XOR of 128 bits (representing integer data) ina
andb
. - _mulx_u32⚠(x86 or x86-64) and
bmi2
Unsigned multiply without affecting flags. - _pdep_u32⚠(x86 or x86-64) and
bmi2
Scatter contiguous low order bits ofa
to the result at the positions specified by themask
. - _pext_u32⚠(x86 or x86-64) and
bmi2
Gathers the bits ofx
specified by themask
into the contiguous low order bit positions of the result. - _popcnt32⚠(x86 or x86-64) and
popcnt
Counts the bits that are set. - _rdrand16_step⚠(x86 or x86-64) and
rdrand
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. - _rdrand32_step⚠(x86 or x86-64) and
rdrand
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. - _rdseed16_step⚠(x86 or x86-64) and
rdseed
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. - _rdseed32_step⚠(x86 or x86-64) and
rdseed
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. - _rdtsc⚠x86 or x86-64Reads the current value of the processor’s time-stamp counter.
- _subborrow_u32⚠x86 or x86-64Adds 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). - _t1mskc_u32⚠(x86 or x86-64) and
tbm
Clears all bits below the least significant zero ofx
and sets all other bits. - _t1mskc_u64⚠(x86 or x86-64) and
tbm
Clears all bits below the least significant zero ofx
and sets all other bits. - _tzcnt_u32⚠(x86 or x86-64) and
bmi1
Counts the number of trailing least significant zero bits. - _tzmsk_u32⚠(x86 or x86-64) and
tbm
Sets all bits below the least significant one ofx
and clears all other bits. - _tzmsk_u64⚠(x86 or x86-64) and
tbm
Sets all bits below the least significant one ofx
and clears all other bits. - _xgetbv⚠(x86 or x86-64) and
xsave
Reads the contents of the extended control registerXCR
specified inxcr_no
. - _xrstor⚠(x86 or x86-64) and
xsave
Performs a full or partial restore of the enabled processor states using the state information stored in memory atmem_addr
. - _xrstors⚠(x86 or x86-64) and
xsave,xsaves
Performs a full or partial restore of the enabled processor states using the state information stored in memory atmem_addr
. - _xsave⚠(x86 or x86-64) and
xsave
Performs a full or partial save of the enabled processor states to memory atmem_addr
. - _xsavec⚠(x86 or x86-64) and
xsave,xsavec
Performs a full or partial save of the enabled processor states to memory atmem_addr
. - _xsaveopt⚠(x86 or x86-64) and
xsave,xsaveopt
Performs a full or partial save of the enabled processor states to memory atmem_addr
. - _xsaves⚠(x86 or x86-64) and
xsave,xsaves
Performs a full or partial save of the enabled processor states to memory atmem_addr
- _xsetbv⚠(x86 or x86-64) and
xsave
Copies 64-bits fromval
to the extended control register (XCR
) specified bya
. - A utility function for creating masks to use with Intel shuffle and permute intrinsics.
- Add 32-bit masks in a and b, and store the result in k.
- Add 64-bit masks in a and b, and store the result in k.
- Compute the bitwise AND of 16-bit masks a and b, and store the result in k.
- Compute the bitwise AND of 32-bit masks a and b, and store the result in k.
- Compute the bitwise AND of 64-bit masks a and b, and store the result in k.
- Compute the bitwise NOT of 16-bit masks a and then AND with b, and store the result in k.
- Compute the bitwise NOT of 32-bit masks a and then AND with b, and store the result in k.
- Compute the bitwise NOT of 64-bit masks a and then AND with b, and store the result in k.
- Compute the bitwise NOT of 16-bit mask a, and store the result in k.
- Compute the bitwise NOT of 32-bit mask a, and store the result in k.
- Compute the bitwise NOT of 64-bit mask a, and store the result in k.
- Compute the bitwise OR of 16-bit masks a and b, and store the result in k.
- Compute the bitwise OR of 32-bit masks a and b, and store the result in k.
- Compute the bitwise OR of 64-bit masks a and b, and store the result in k.
- Compute the bitwise XNOR of 16-bit masks a and b, and store the result in k.
- Compute the bitwise XNOR of 32-bit masks a and b, and store the result in k.
- Compute the bitwise XNOR of 64-bit masks a and b, and store the result in k.
- Compute the bitwise XOR of 16-bit masks a and b, and store the result in k.
- Compute the bitwise XOR of 32-bit masks a and b, and store the result in k.
- Compute the bitwise XOR of 64-bit masks a and b, and store the result in k.
- Load 32-bit mask from memory into k.
- Load 64-bit mask from memory into k.
- Compute the absolute value of packed signed 64-bit integers in a, and store the unsigned results in dst.
- Performs one round of an AES decryption flow on each 128-bit word (state) in
a
using the corresponding 128-bit word (key) inround_key
. - Performs the last round of an AES decryption flow on each 128-bit word (state) in
a
using the corresponding 128-bit word (key) inround_key
. - Performs one round of an AES encryption flow on each 128-bit word (state) in
a
using the corresponding 128-bit word (key) inround_key
. - Performs the last round of an AES encryption flow on each 128-bit word (state) in
a
using the corresponding 128-bit word (key) inround_key
. - 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.
- 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.
- Considers the input
b
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. - Broadcast the 4 packed single-precision (32-bit) floating-point elements from a to all elements of dst.
- Broadcast the 4 packed 32-bit integers from a to all elements of dst.
- Broadcast the low 8-bits from input mask k to all 64-bit elements of dst.
- Broadcast the low 16-bits from input mask k to all 32-bit elements of dst.
- Performs a carry-less multiplication of two 64-bit polynomials over the finite field GF(2) - in each of the 2 128-bit lanes.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- Compare packed signed 8-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed 32-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed 64-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for equality, and store the results in mask vector k.
- Compare packed signed 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed signed 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed signed 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for greater-than-or-equal, and store the results in mask vector k.
- Compare packed signed 8-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed signed 32-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed signed 64-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for greater-than, and store the results in mask vector k.
- Compare packed signed 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed signed 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed signed 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for less-than-or-equal, and store the results in mask vector k.
- Compare packed signed 8-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed signed 32-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed signed 64-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for less-than, and store the results in mask vector k.
- Compare packed signed 8-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed signed 16-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed 32-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed signed 64-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed unsigned 8-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed unsigned 16-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed unsigned 32-bit integers in a and b for not-equal, and store the results in mask vector k.
- Compare packed unsigned 64-bit integers in a and b for not-equal, and store the results in mask vector k.
- 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.
- 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.
- Convert packed 16-bit integers in a to packed 8-bit integers with truncation, and store the results in dst.
- Convert packed 32-bit integers in a to packed 8-bit integers with truncation, and store the results in dst.
- Convert packed 32-bit integers in a to packed 16-bit integers with truncation, and store the results in dst.
- Convert packed 64-bit integers in a to packed 8-bit integers with truncation, and store the results in dst.
- Convert packed 64-bit integers in a to packed 16-bit integers with truncation, and store the results in dst.
- Convert packed 64-bit integers in a to packed 32-bit integers with truncation, and store the results in dst.
- Convert packed unsigned 32-bit integers in a to packed double-precision (64-bit) floating-point elements, and store the results in dst.
- 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
- 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
- Convert packed double-precision (64-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst.
- Convert packed single-precision (32-bit) floating-point elements in a to packed unsigned 32-bit integers, and store the results in dst.
- Convert packed signed 16-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst.
- Convert packed signed 32-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst.
- Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst.
- Convert packed signed 64-bit integers in a to packed 8-bit integers with signed saturation, and store the results in dst.
- Convert packed signed 64-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst.
- Convert packed signed 64-bit integers in a to packed 32-bit integers with signed saturation, and store the results in dst.
- 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.
- 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.
- Convert packed unsigned 16-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst.
- Convert packed unsigned 32-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst.
- Convert packed unsigned 32-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst.
- Convert packed unsigned 64-bit integers in a to packed unsigned 8-bit integers with unsigned saturation, and store the results in dst.
- Convert packed unsigned 64-bit integers in a to packed unsigned 16-bit integers with unsigned saturation, and store the results in dst.
- Convert packed unsigned 64-bit integers in a to packed unsigned 32-bit integers with unsigned saturation, and store the results in dst.
- 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.
- 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
- 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.
- 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.
- 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.
- 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.
- 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.
- Extract 128 bits (composed of 4 packed 32-bit integers) from a, selected with IMM1, and store the result in dst.
- 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.
- 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.
- 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.
- 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.
- 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 - 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
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- Counts the number of leading zero bits in each packed 32-bit integer in a, and store the results in dst.
- Counts the number of leading zero bits in each packed 64-bit integer in a, and store the results in dst.
- Multiply packed unsigned 52-bit integers in each 64-bit element of
b
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
. - Multiply packed unsigned 52-bit integers in each 64-bit element of
b
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
. - 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).
- 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).
- 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).
- 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).
- 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)
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- Considers the input
b
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. - Blend packed 8-bit integers from a and b using control mask k, and store the results in dst.
- Blend packed 16-bit integers from a and b using control mask k, and store the results in dst.
- Blend packed 32-bit integers from a and b using control mask k, and store the results in dst.
- Blend packed 64-bit integers from a and b using control mask k, and store the results in dst.
- Blend packed double-precision (64-bit) floating-point elements from a and b using control mask k, and store the results in dst.
- Blend packed single-precision (32-bit) floating-point elements from a and b using control mask k, and store the results in dst.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- Contiguously store the active 8-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr.
- Contiguously store the active 16-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr.
- Contiguously store the active 32-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr.
- Contiguously store the active 64-bit integers in a (those with their respective bit set in writemask k) to unaligned memory at base_addr.
- 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.
- 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.
- 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.
- 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.
- 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 - 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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
- 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
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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 - 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).
- 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.
- 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).
- 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).
- 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.
- 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.
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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).
- 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).
- 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.
- 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.
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- 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).
- 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).
- 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
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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(x86 or x86-64) and
avx512f,avx512bw,avx512vbmi2,avx512vl,avx
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). - 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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 - 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 - 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.
- 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.
- 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.
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- For each packed 8-bit integer maps the value to the number of logical 1 bits.
- For each packed 16-bit integer maps the value to the number of logical 1 bits.
- For each packed 32-bit integer maps the value to the number of logical 1 bits.
- For each packed 64-bit integer maps the value to the number of logical 1 bits.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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 - 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 - 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- Store packed 8-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary.
- Store packed 16-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary.
- Store packed 32-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary.
- Store packed 64-bit integers from a into memory using writemask k. mem_addr does not need to be aligned on any particular boundary.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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.
- 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 - 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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
- 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
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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 - 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).
- 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).
- Convert packed signed 32-bit integers in a to packed 16-bit integers with signed saturation, and store the results in dst.
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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).
- 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.
- 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_ps⚠Experimental