leavelet/cutlass
1
0
Fork 0
Code Issues Pull Requests Actions 25 Packages Projects Releases Wiki Activity

v4.2 release. (#2587)

Browse Source 
* Fix default cluster callback values to 1 to avoid profiler failure when these values are not set in command line.

* v4.2 release.
This commit is contained in:
Junkai-Wu
2025-08-22 18:11:24 -04:00
committed by GitHub
parent 11cad1f67b
commit a49a78ffef
351 changed files with 28182 additions and 2032 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/70_blackwell_gemm/70_blackwell_fp16_gemm.cu
+3 -2
View File
@@ -454,11 +454,12 @@ int main(int argc, char const **args) {
CUDA_CHECK(cudaGetDevice(&current_device_id));
CUDA_CHECK(cudaGetDeviceProperties(&props, current_device_id));
cudaError_t error = cudaGetDeviceProperties(&props, 0);
if (props.major != 10 || props.minor != 0) {
std::cerr << "This example requires a GPU with compute capability 100a)." << std::endl;
return 0;
}
}
//
// Parse options
//
examples/70_blackwell_gemm/70_blackwell_fp8_gemm.cu
+2 -2
View File
@@ -640,11 +640,11 @@ int main(int argc, char const **args) {
CUDA_CHECK(cudaGetDevice(&current_device_id));
CUDA_CHECK(cudaGetDeviceProperties(&props, current_device_id));
cudaError_t error = cudaGetDeviceProperties(&props, 0);
if (props.major != 10 || props.minor != 0) {
std::cerr << "This example requires a GPU with compute capability 100a)." << std::endl;
return 0;
}
}
//
// Parse options
examples/70_blackwell_gemm/CMakeLists.txt
+1 -1
View File
@@ -33,7 +33,7 @@ set(TEST_SWIZZLE_2 --swizzle=2)
set(TEST_SWIZZLE_5 --swizzle=5)
set(TEST_SWIZZLE_5_UNEVEN --swizzle=5 --m=4096 --n=16384)
if(NOT CUTLASS_NVCC_ARCHS STREQUAL "100")
if(CUTLASS_NVCC_ARCHS STREQUAL "100a" OR CUTLASS_NVCC_ARCHS STREQUAL "100f" OR CUTLASS_NVCC_ARCHS STREQUAL "101a" OR CUTLASS_NVCC_ARCHS STREQUAL "101f" OR CUTLASS_NVCC_ARCHS STREQUAL "103a" OR CUTLASS_NVCC_ARCHS STREQUAL "103f")
cutlass_example_add_executable(
70_blackwell_fp16_gemm
70_blackwell_fp16_gemm.cu