CUDA_gemm
CUDA_gemm copied to clipboard
fixed the boundary condition
See https://github.com/Cjkkkk/CUDA_gemm/issues/6.