Cutlass
CUDA Templates for Linear Algebra Subroutines and Solvers
|
Partial specialization for launching the GEMM kernel with or without launch bounds.
#include <gemm.h>
Public Member Functions | |
Launch (typename Gemm::Params params, dim3 grid, dim3 block, cudaStream_t stream=0) | |
|
inline |