leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 28 Packages Projects Releases Wiki Activity

Release v4.0.0 (#2294)

Browse Source
This commit is contained in:
Kihiro Bando
2025-05-13 15:55:29 -04:00
committed by GitHub
parent ad7b2f5e84
commit f115c3f854
299 changed files with 51495 additions and 4413 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/cute/tutorial/hopper/wgmma_sm90.cu
+4 -4
View File
@@ -506,13 +506,13 @@ int main(int argc, char** argv)
return -1;
}
if (props.major < 8) {
std::cout << "This example requires an Ampere GPU or newer (CC >= 80)" << std::endl;
if (props.major != 9) {
std::cout << "This example requires NVIDIA's Hopper Architecture GPU with compute capability 90a" << std::endl;
// Return 0 so tests pass if run on unsupported architectures or CUDA Toolkits.
return 0;
}
#if defined(CUTLASS_ARCH_MMA_SM90A_SUPPORTED)
#if defined(CUTLASS_ARCH_MMA_SM90_SUPPORTED)
int m = 5120;
if (argc >= 2)
@@ -604,7 +604,7 @@ int main(int argc, char** argv)
printf("CUTE_GEMM: [%6.1f]GFlop/s (%6.4f)ms\n", gflops / cute_time, cute_time*1000);
#else
std::cout << "CUTLASS_ARCH_MMA_SM90A_SUPPORTED must be enabled, but it is not. Test is waived \n" << std::endl;
std::cout << "CUTLASS_ARCH_MMA_SM90_SUPPORTED must be enabled, but it is not. Test is waived \n" << std::endl;
#endif
return 0;
examples/cute/tutorial/hopper/wgmma_tma_sm90.cu
+2 -2
View File
@@ -461,7 +461,7 @@ int main(int argc, char** argv)
return 0;
}
#if defined(CUTLASS_ARCH_MMA_SM90A_SUPPORTED)
#if defined(CUTLASS_ARCH_MMA_SM90_SUPPORTED)
int m = 512;
if (argc >= 2)
@@ -553,7 +553,7 @@ int main(int argc, char** argv)
printf("CUTE_GEMM: [%6.1f]GFlop/s (%6.4f)ms\n", gflops / cute_time, cute_time*1000);
#else
std::cout << "CUTLASS_ARCH_MMA_SM90A_SUPPORTED must be enabled, but it is not. Test is waived \n" << std::endl;
std::cout << "CUTLASS_ARCH_MMA_SM90_SUPPORTED must be enabled, but it is not. Test is waived \n" << std::endl;
#endif
return 0;