Intel® C++ Compiler Classic Developer Guide and Reference

ID 767249
Date 12/16/2022
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

Intrinsics for Reciprocal Operations (512-bit)

The prototypes for Intel® Advanced Vector Extensions 512 (Intel® AVX-512) intrinsics are located in the zmmintrin.h header file.

To use these intrinsics, include the immintrin.h file as follows:

#include <immintrin.h>


Intrinsic Name

Operation

Corresponding
Intel® AVX-512 Instruction

_mm512_rcp14_pd, _mm512_mask_rcp14_pd, _mm512_maskz_rcp14_pd

Computes the approximate reciprocal of packed float64 elements.

VRCP14PD

_mm512_rcp14_ps, _mm512_mask_rcp14_ps, _mm512_maskz_rcp14_ps

Computes the approximate reciprocal of packed float32 elements.

VRCP14PS

_mm_rcp14_sd, _mm_mask_rcp14_sd, _mm_maskz_rcp14_sd

Computes the approximate reciprocal of scalar float64 elements.

VRCP14SD

_mm_rcp14_ss, _mm_mask_rcp14_ss, _mm_maskz_rcp14_ss

Computes the approximate reciprocal of scalar float32 elements.

VRCP14SS

_mm512_rcp28_pd, _mm512_mask_rcp28_pd, _mm512_maskz_rcp28_pd

_mm512_rcp28_round_pd, _mm512_mask_rcp28_round_pd, _mm512_maskz_rcp28_round_pd

Computes the approximate reciprocal of packed float64 elements with bounded relative error.

VRCP28PD

_mm_rcp28_sd, _mm_mask_rcp28_sd, _mm_maskz_rcp28_sd

_mm_rcp28_round_sd, _mm_mask_rcp28_round_sd, _mm_maskz_rcp28_round_sd

Computes the approximate reciprocal of scalar float64 elements with bounded relative error.

VRCP28SD

_mm512_rcp28_ps, _mm512_mask_rcp28_ps, _mm512_maskz_rcp28_ps

_mm512_rcp28_round_ps, _mm512_mask_rcp28_round_ps, _mm512_maskz_rcp28_round_ps

Computes the approximate reciprocal of packed float32 elements with bounded relative error.

VRCP28PS

_mm_rcp28_ss, _mm_mask_rcp28_ss, _mm_maskz_rcp28_ss

_mm_rcp28_round_ss, _mm_mask_rcp28_round_ss, _mm_maskz_rcp28_round_ss

Computes the approximate reciprocal of scalar float32 elements with bounded relative error.

VRCP28SS

_mm512_recip_pd, _mm512_mask_recip_pd

Computes the approximate reciprocal of packed float64 elements.

None.

_mm512_recip_ps, _mm512_mask_recip_ps

Computes the approximate reciprocal of packed float32 elements.

None.


variable definition
k

writemask used as a selector

a

first source vector element

b

second source vector element

src

source element to use based on writemask result


_mm512_rcp14_pd

extern __m512d __cdecl _mm512_rcp14_pd(__m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm512_mask_rcp14_pd

extern __m512d __cdecl _mm512_mask_rcp14_pd(__m512d src, __mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm512_maskz_rcp14_pd

extern __m512d __cdecl _mm512_maskz_rcp14_pd(__mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-14).



_mm512_rcp14_ps

extern __m512 __cdecl _mm512_rcp14_ps(__m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm512_mask_rcp14_ps

extern __m512 __cdecl _mm512_mask_rcp14_ps(__m512 src, __mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm512_maskz_rcp14_ps

extern __m512 __cdecl _mm512_maskz_rcp14_ps(__mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-14).



_mm_rcp14_sd

extern __m128d __cdecl _mm_rcp14_sd(__m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element, and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm_mask_rcp14_sd

extern __m128d __cdecl _mm_mask_rcp14_sd(__m128d src, __mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm_maskz_rcp14_sd

extern __m128d __cdecl _mm_maskz_rcp14_sd(__mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-14).



_mm_rcp14_ss

extern __m128 __cdecl _mm_rcp14_ss(__m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element, and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm_mask_rcp14_ss

extern __m128 __cdecl _mm_mask_rcp14_ss(__m128 src, __mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-14).


_mm_maskz_rcp14_ss

extern __m128 __cdecl _mm_maskz_rcp14_ss(__mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-14).



_mm512_rcp28_round_pd

extern __m512d __cdecl _mm512_rcp28_round_pd(__m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_round_pd

extern __m512d __cdecl _mm512_mask_rcp28_round_pd(__m512d src, __mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_round_pd

extern __m512d __cdecl _mm512_maskz_rcp28_round_pd(__mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_pd

extern __m512d __cdecl _mm512_rcp28_pd(__m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_pd

extern __m512d __cdecl _mm512_mask_rcp28_pd(__m512d src, __mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_pd

extern __m512d __cdecl _mm512_maskz_rcp28_pd(__mmask8 k, __m512d a);

Computes the approximate reciprocal of packed float64 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_round_ps

extern __m512 __cdecl _mm512_rcp28_round_ps(__m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_round_ps

extern __m512 __cdecl _mm512_mask_rcp28_round_ps(__m512 src, __mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_round_ps

extern __m512 __cdecl _mm512_maskz_rcp28_round_ps(__mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_ps

extern __m512 __cdecl _mm512_rcp28_ps(__m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_ps

extern __m512 __cdecl _mm512_mask_rcp28_ps(__m512 src, __mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_ps

extern __m512 __cdecl _mm512_maskz_rcp28_ps(__mmask16 k, __m512 a);

Computes the approximate reciprocal of packed float32 elements in a, and stores the result using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_round_sd

extern __m128d __cdecl _mm512_rcp28_round_sd(__m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element, and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_round_sd

extern __m128d __cdecl _mm512_mask_rcp28_round_sd(__m128d src, __mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_round_sd

extern __m128d __cdecl _mm512_maskz_rcp28_round_sd(__mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_round_sd

extern __m128d __cdecl _mm512_rcp28_round_sd(__m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element, and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_round_sd

extern __m128d __cdecl _mm512_mask_rcp28_round_sd(__m128d src, __mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_round_sd

extern __m128d __cdecl _mm512_maskz_rcp28_round_sd(__mmask8 k, __m128d a, __m128d b);

Computes the approximate reciprocal of lower float64 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper element from a to upper destination element.

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_round_ss

extern __m128 __cdecl _mm512_rcp28_round_ss(__m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element, and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_round_ss

extern __m128 __cdecl _mm512_mask_rcp28_round_ss(__m128 src, __mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_round_ss

extern __m128 __cdecl _mm512_maskz_rcp28_round_ss(__mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_rcp28_ss

extern __m128 __cdecl _mm512_rcp28_ss(__m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element, and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_mask_rcp28_ss

extern __m128 __cdecl _mm512_mask_rcp28_ss(__m128 src, __mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).


_mm512_maskz_rcp28_ss

extern __m128 __cdecl _mm512_maskz_rcp28_ss(__mmask8 k, __m128 a, __m128 b);

Computes approximate reciprocal of lower float32 element in b, stores the result in lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies upper three packed elements from a to upper destination elements.

NOTE:

The maximum relative error for this approximation is less than 2(-28).



_mm512_recip_pd

extern __m512d __cdecl _mm512_recip_pd(__m512d a);

Computes approximate reciprocal of float64 elements in a, and stores the result.


_mm512_mask_recip_pd

extern __m512d __cdecl _mm512_mask_recip_pd(__m512d src, __mmask8 k, __m512d a);

Computes approximate reciprocal of float64 elements in a, and stores the result using writemask k (the element is copied from src when mask bit 0 is not set).



_mm512_recip_ps

extern __m512 __cdecl _mm512_recip_ps(__m512 a);

Computes approximate reciprocal of float32 elements in a, and stores the result.


_mm512_mask_recip_ps

extern __m512 __cdecl _mm512_mask_recip_ps(__m512 src, __mmask16 k, __m512 a);

Computes approximate reciprocal of float32 elements in a, and stores the result using writemask k (the element is copied from src when mask bit 0 is not set).