Intel® C++ Compiler 16.0 User and Reference Guide
Determines the maximum of float32 vectors. Corresponding instruction is VGMAXPS. This intrinsic only applies to Intel® Many Integrated Core Architecture (Intel® MIC Architecture).
Without Mask extern _m512 __cdecl _mm512_gmax_ps(_m512 v2, _m512 v3); |
With Mask extern _m512 __cdecl _mm512_gmask_max_ps(_m512 v1_old, __mmask16 k1, _m512 v2, _m512 v3); |
v2 |
float32 vector to determine the maximum values (first source operand) |
v3 |
float32 vector to determine the maximum values (second source operand) |
v1_old |
Source vector that retains old values of the destination vector; the resulting vector gets corresponding elements from v1_old for zero mask bits |
k1 |
Writemask; only those elements of the source vectors with corresponding bit set to '1' in the k1 mask are computed and stored in the result; elements in the result vector corresponding to zero bit in k1 are copied from corresponding elements of vector v1_old |
Determines the maximum value of each pair of corresponding elements in float32 vector v2 and float32 vector v3.
FpMax() returns the larger of the two float32 arguments, following IEEE in general, and specifically following this from the DX10 spec:
dest = src0 >= src1 ? src0 : src1
The >= is used instead of >, so that if min(x,y) = x then max(x,y) = y.
NaN has special handling: If one source operand is NaN, then the other source operand is returned (choice made per-component). If both are NaN, then the quietized NaN from the first source (here v2) is returned. This conforms to new IEEE 754R rules.
The masked variant has one additional argument: k1. Only those elements in the source registers with the corresponding bit set in vector mask k1 are used for computing.
Returns the result of the operation.