Displaying 1 result from an estimated 1 matches for "zmm16".
Did you mean:
zmm1
2017 Jul 01
2
KNL Assembly Code for Matrix Multiplication
...j
>>>>> changes whats the need of this step?*
>>>>> * vpsllq zmm15, zmm15, 2 ; **dont understand the need for this step*
>>>>> * vpaddq zmm14, zmm14, zmm15 ; **dont understand the need for this
>>>>> step*
>>>>> * vpmuludq zmm16, zmm9, zmm2 ; **here same as before the lower 8
>>>>> elements of B indexes are computed as Zmm16=[0,4000,8000,.......28000]*
>>>>> * vpsrlq zmm17, zmm9, 32 **; **dont understand the need for this
>>>>> step*
>>>>> * vpmuludq zmm17, zmm1...