7cfa87b5ac
siboehm/SGEMM_CUDA (19 files):
siboehm_sgemm.cu, siboehm_runner.cu, siboehm_runner.cuh, siboehm_kernels.cuh
siboehm_cuBLAS_sgemm.cu, siboehm_simplest_kernel.cu, siboehm_CMakeLists.txt
siboehm_{1_naive..12_kernel_double_buffering}.cuh
wangzyon/NVIDIA_SGEMM_PRACTICE (12 files):
wangzyon_sgemm.cu, wangzyon_utils.cu, wangzyon_utils.cuh, wangzyon_kernel.cuh
wangzyon_CMakeLists.txt, wangzyon_kernel_{1..7}.cuh
edtallison/sgemm-cuda (19 files):
edtallison_sgemm.cu, edtallison_runner.cu, edtallison_runner.cuh
edtallison_kernels.cuh, edtallison_cuBLAS_sgemm.cu, edtallison_simplest_kernel.cu
edtallison_CMakeLists.txt, edtallison_{01_naive..12_kernel_double_buffering}.cuh
cat_files/ total: 25 → 75 files
24 lines
796 B
Plaintext
24 lines
796 B
Plaintext
#pragma once
|
|
|
|
#include <cassert>
|
|
#include <cstdio>
|
|
#include <cstdlib>
|
|
#include <cublas_v2.h>
|
|
#include <cuda_runtime.h>
|
|
|
|
template <const uint BLOCKSIZE>
|
|
__global__ void sgemm_global_mem_coalesce(int M, int N, int K, float alpha,
|
|
const float *A, const float *B,
|
|
float beta, float *C) {
|
|
const int cRow = blockIdx.x * BLOCKSIZE + (threadIdx.x / BLOCKSIZE);
|
|
const int cCol = blockIdx.y * BLOCKSIZE + (threadIdx.x % BLOCKSIZE);
|
|
|
|
// if statement is necessary to make things work under tile quantization
|
|
if (cRow < M && cCol < N) {
|
|
float tmp = 0.0;
|
|
for (int i = 0; i < K; ++i) {
|
|
tmp += A[cRow * K + i] * B[i * N + cCol];
|
|
}
|
|
C[cRow * N + cCol] = alpha * tmp + beta * C[cRow * N + cCol];
|
|
}
|
|
} |