We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent c29d5d4 commit 7179053Copy full SHA for 7179053
ggml/src/ggml-cuda/common.cuh
@@ -316,7 +316,7 @@ static bool ampere_mma_available(const int cc) {
316
}
317
318
static bool cp_async_available(const int cc) {
319
- return cc < GGML_CUDA_CC_OFFSET_AMD && ggml_cuda_highest_compiled_arch(cc) >= GGML_CUDA_CC_AMPERE;
+ return GGML_CUDA_CC_IS_NVIDIA(cc) && ggml_cuda_highest_compiled_arch(cc) >= GGML_CUDA_CC_AMPERE;
320
321
322
static constexpr __device__ int ggml_cuda_get_physical_warp_size() {
0 commit comments