std1.0.0-beta-7 SafeC docs

std :: ml::gpu_rocm

Function std::rocm_attention_use_mfma

#include <std/ml/gpu_rocm.h> source ยท definition

int rocm_attention_use_mfma(int on)

on = 0 makes both ops use the CUDA-core kernels even on matrix-core GPUs (for comparison and testing); on = 1 restores the default. Returns 1 if the matrix-core kernels are in use after the call.