perf(gemm2): 1:1 HIP a4w4 gemm2 port + small-M latency optimization to HIP parity#660
Closed
fsx950223 wants to merge 7 commits into
Closed
perf(gemm2): 1:1 HIP a4w4 gemm2 port + small-M latency optimization to HIP parity#660fsx950223 wants to merge 7 commits into
fsx950223 wants to merge 7 commits into