adding blackwell (#2143)

This commit is contained in:
Mohamed Mekkouri
2025-03-18 03:20:40 +01:00
committed by GitHub
parent bd03b22f64
commit 3fe62887d8

View File

@ -180,6 +180,11 @@ $ make test_unit_gemm_warp -j
To minimize compilation time, specific GPU architectures can be enabled via the CMake command,
selected by [CUDA Compute Capability.](https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#compute-capabilities)
**NVIDIA Blackwell Architecture.**
```bash
$ cmake .. -DCUTLASS_NVCC_ARCHS=100a # compiles for NVIDIA Blackwell GPU architecture
```
**NVIDIA Hopper Architecture.**
```bash
$ cmake .. -DCUTLASS_NVCC_ARCHS=90a # compiles for NVIDIA Hopper GPU architecture