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A newer version of this document is available.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 Determining Minimum and Maximum FP Values

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_max_round_pd_mm512_mask_max_round_pd_mm512_maskz_max_round_pd

_mm512_max_pd_mm512_mask_max_pd_mm512_maskz_max_pd

Calculate maximum of rounded float64 values.

VMAXPD

_mm512_max_round_ps_mm512_mask_max_round_ps_mm512_maskz_max_round_ps

_mm512_max_ps_mm512_mask_max_ps_mm512_maskz_max_ps

Calculate maximum of rounded float32 values.

VMAXPS

_mm_mask_max_sd_mm_maskz_max_sd

_mm_max_round_sd, _mm_mask_max_round_sd, _mm_maskz_max_round_sd

Calculate maximum of scalar float64 values.

VMAXSD

_mm_mask_max_ss_mm_maskz_max_ss

_mm_max_round_ss, _mm_mask_max_round_ss, _mm_maskz_max_round_ss

Calculate maximum of scalar float32 values.

VMAXSS

_mm512_min_pd, _mm512_mask_min_pd, _mm512_maskz_min_pd

_mm512_min_round_pd, _mm512_mask_min_round_pd, _mm512_maskz_min_round_pd

Calculate minimum of packed float64 values.

VMINPD

_mm512_min_ps, _mm512_mask_min_ps, _mm512_maskz_min_ps

_mm512_min_round_ps, _mm512_mask_min_round_ps, _mm512_maskz_min_round_ps

Calculate minimum of packed float32 values.

VMINPS

_mm_mask_min_sd_mm_maskz_min_sd

_mm_min_round_sd, _mm_mask_min_round_sd, _mm_maskz_min_round_sd

Calculate minimum of scalar float64 values.

VMINSD

_mm_mask_min_ss_mm_maskz_min_ss

_mm_min_round_ss, _mm_mask_min_round_ss, _mm_maskz_min_round_ss

Calculate minimum of scalar float32 values.

VMINSS


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

round

Rounding control values; these can be one of the following (along with the sae suppress all exceptions flag):

  • _MM_FROUND_TO_NEAREST_INT - rounds to nearest even
  • _MM_FROUND_TO_NEG_INF - rounds to negative infinity
  • _MM_FROUND_TO_POS_INF - rounds to positive infinity
  • _MM_FROUND_TO_ZERO - rounds to zero
  • _MM_FROUND_CUR_DIRECTION - rounds using default from MXCSR register


_mm512_max_pd

extern __m512d __cdecl _mm512_max_pd(__m512d a, __m512d b);

Compares packed float64 elements in a and b, and stores packed maximum values.




_mm512_mask_max_pd

extern __m512d __cdecl _mm512_mask_max_pd(__m512d src, __mmask8 k, __m512d a, __m512d b);

Compares packed float64 elements in a and b, and stores packed maximum values using writemask k (elements are copied from src when the corresponding mask bit is not set).




_mm512_maskz_max_pd

extern __m512d __cdecl _mm512_maskz_max_pd(__mmask8 k, __m512d a, __m512d b);

Compares packed float64 elements in a and b, and stores packed maximum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).




_mm512_max_round_pd

extern __m512d __cdecl _mm512_max_round_pd(__m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed maximum values.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_mask_max_round_pd

extern __m512d __cdecl _mm512_mask_max_round_pd(__m512d src, __mmask8 k, __m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed maximum values using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_maskz_max_round_pd

extern __m512d __cdecl _mm512_maskz_max_round_pd(__mmask8 k, __m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed maximum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_max_ps

extern __m512 __cdecl _mm512_max_ps(__m512 a, __m512 b);

Compares packed float32 elements in a and b, and stores packed maximum values.




_mm512_mask_max_ps

extern __m512 __cdecl _mm512_mask_max_ps(__m512 src, __mmask16 k, __m512 a, __m512 b);

Compares packed float32 elements in a and b, and stores packed maximum values using writemask k (elements are copied from src when the corresponding mask bit is not set).




_mm512_maskz_max_ps

extern __m512 __cdecl _mm512_maskz_max_ps(__mmask16 k, __m512 a, __m512 b);

Compares packed float32 elements in a and b, and stores packed maximum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).




_mm512_max_round_ps

extern __m512 __cdecl _mm512_max_round_ps(__m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed maximum values.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_mask_max_round_ps

extern __m512 __cdecl _mm512_mask_max_round_ps(__m512 src, __mmask16 k, __m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed maximum values using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_maskz_max_round_ps

extern __m512 __cdecl _mm512_maskz_max_round_ps(__mmask16 k, __m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed maximum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_max_sd

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

Compares the lower float64 elements in a and b, stores the maximum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_maskz_max_sd

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

Compares the lower float64 elements in a and b, stores the maximum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_max_round_sd

extern __m128d __cdecl _mm_max_round_sd(__m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the maximum value in the lower destination element, and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_max_round_sd

extern __m128d __cdecl _mm_mask_max_round_sd(__m128d src, __mmask8 k, __m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the maximum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_maskz_max_round_sd

extern __m128d __cdecl _mm_maskz_max_round_sd(__mmask8 k, __m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the maximum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_max_ss

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

Compares the lower float32 elements in a and b, stores the maximum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_maskz_max_ss

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

Compares the lower float32 elements in a and b, stores the maximum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_max_round_ss

extern __m128 __cdecl _mm_max_round_ss(__m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the maximum value in the lower destination element, and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_max_round_ss

extern __m128 __cdecl _mm_mask_max_round_ss(__m128 src, __mmask8 k, __m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the maximum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_maskz_max_round_ss

extern __m128 __cdecl _mm_maskz_max_round_ss(__mmask8 k, __m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the maximum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_min_pd

extern __m512d __cdecl _mm512_min_pd(__m512d a, __m512d b);

Compares packed float64 elements in a and b, and stores packed minimum values.




_mm512_mask_min_pd

extern __m512d __cdecl _mm512_mask_min_pd(__m512d src, __mmask8 k, __m512d a,__m512d b);

Compares packed float64 elements in a and b, and stores packed minimum values using writemask k (elements are copied from src when the corresponding mask bit is not set).




_mm512_maskz_min_pd

extern __m512d __cdecl _mm512_maskz_min_pd(__mmask8 k, __m512d a, __m512d b);

Compares packed float64 elements in a and b, and store packed minimum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).




_mm512_min_round_pd

extern __m512d __cdecl _mm512_min_round_pd(__m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed minimum values.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_mask_min_round_pd

extern __m512d __cdecl _mm512_mask_min_round_pd(__m512d src, __mmask8 k, __m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed minimum values using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_maskz_min_round_pd

extern __m512d __cdecl _mm512_maskz_min_round_pd(__mmask8 k, __m512d a, __m512d b, int round);

Compares packed float64 elements in a and b, and stores packed minimum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_min_ps

extern __m512 __cdecl _mm512_min_ps(__m512 a, __m512 b);

Compares packed float32 elements in a and b, and stores packed minimum values.



_mm512_mask_min_ps

extern __m512 __cdecl _mm512_mask_min_ps(__m512 src, __mmask16 k, __m512 a, __m512 b);

Compares packed float32 elements in a and b, and stores packed minimum values using writemask k (elements are copied from src when the corresponding mask bit is not set).



_mm512_maskz_min_ps

extern __m512 __cdecl _mm512_maskz_min_ps(__mmask16 k, __m512 a, __m512 b;

Compares packed float32 elements in a and b, and store packed minimum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).




_mm512_min_round_ps

extern __m512 __cdecl _mm512_min_round_ps( __m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed minimum values.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_mask_min_round_ps

extern __m512 __cdecl _mm512_mask_min_round_ps(__m512 src, __mmask16 k, __m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed minimum values using writemask k (elements are copied from src when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm512_maskz_min_round_ps

extern __m512 __cdecl _mm512_maskz_min_round_ps(__mmask16 k, __m512 a, __m512 b, int round);

Compares packed float32 elements in a and b, and stores packed minimum values using zeromask k (elements are zeroed out when the corresponding mask bit is not set).

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_min_sd

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

Compares the lower float64 elements in a and b, stores the minimum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_maskz_min_sd

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

Compares the lower float64 elements in a and b, stores the minimum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_min_round_sd

extern __m128d __cdecl _mm_min_round_sd(__m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the minimum value in the lower destination element, and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_min_round_sd

extern __m128d __cdecl _mm_mask_min_round_sd(__m128d src, __mmask8 k, __m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the minimum value in the lower destination element of using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_maskz_min_round_sd

extern __m128d __cdecl _mm_maskz_min_round_sd(__mmask8 k, __m128d a, __m128d b, int round);

Compares the lower float64 elements in a and b, stores the minimum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_min_ss

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

Compares the lower float32 elements in a and b, stores the minimum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_maskz_min_ss

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

Compares the lower float32 elements in a and b, stores the minimum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.




_mm_min_round_ss

extern __m128 __cdecl _mm_min_round_ss(__m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the minimum value in the lower destination element, and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_mask_min_round_ss

extern __m128 __cdecl _mm_mask_min_round_ss(__m128 src, __mmask8 k, __m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the minimum value in the lower destination element using writemask k (the element is copied from src when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.




_mm_maskz_min_round_ss

extern __m128 __cdecl _mm_maskz_min_round_ss(__mmask8 k, __m128 a, __m128 b, int round);

Compares the lower float32 elements in a and b, stores the minimum value in the lower destination element using zeromask k (the element is zeroed out when mask bit 0 is not set), and copies the upper element from a to the upper destination element.

NOTE:

Pass __MM_FROUND_NO_EXC to round to suppress all exceptions.