Add cutlass support for blackwell fp8 blockwise gemm (#14383)
Signed-off-by: Shu Wang <shuw@nvidia.com>
This commit is contained in:
@@ -59,3 +59,13 @@ struct enable_sm90_only : Kernel {
|
||||
#endif
|
||||
}
|
||||
};
|
||||
|
||||
template <typename Kernel>
|
||||
struct enable_sm100_only : Kernel {
|
||||
template <typename... Args>
|
||||
CUTLASS_DEVICE void operator()(Args&&... args) {
|
||||
#if defined __CUDA_ARCH__ && __CUDA_ARCH__ == 1000
|
||||
Kernel::operator()(std::forward<Args>(args)...);
|
||||
#endif
|
||||
}
|
||||
};
|
||||
|
||||
Reference in New Issue
Block a user