⭐ 欢迎来到虫虫下载站! | 📦 资源下载 📁 资源专辑 ℹ️ 关于我们
⭐ 虫虫下载站

📄 sgemm.cu

📁 Nividia提供的CUDA的BLAS库源码
💻 CU
📖 第 1 页 / 共 2 页
字号:
#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_gld_sw_ta_tb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_gld_sw_na_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY #undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_gld_sw_na_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_gld_sw_ta_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_gld_sw_ta_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           0#include "sgemm.h"}__global__ void sgemm_main_tex_hw_na_nb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 0#define TRANSA            0#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_na_tb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 0#define TRANSA            0#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_ta_nb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 0#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_ta_tb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY   #undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 0#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_na_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 0#define TRANSA            0#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_na_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 0#define TRANSA            0#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_ta_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 0#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_hw_ta_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY   #undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 0#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_na_nb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY #undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_na_tb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_ta_nb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_ta_tb (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   0#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_na_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY #undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_na_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            0#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_ta_nb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            0#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}__global__ void sgemm_main_tex_sw_ta_tb_fulltile (struct cublasSgemmParams parms){#undef  FULL_TILES_ONLY#undef  USE_MIXED_STEPPER#undef  TRANSA#undef  TRANSB#undef  FAST_IMUL#undef  USE_TEX#define FULL_TILES_ONLY   1#define USE_MIXED_STEPPER 1#define TRANSA            1#define TRANSB            1#define FAST_IMUL         0#define USE_TEX           1#include "sgemm.h"}

⌨️ 快捷键说明

复制代码 Ctrl + C
搜索代码 Ctrl + F
全屏模式 F11
切换主题 Ctrl + Shift + D
显示快捷键 ?
增大字号 Ctrl + =
减小字号 Ctrl + -