Originally published at: https://developer.nvidia.com/blog/just-released-nvidia-cusparselt-0-6/
NVIDIA cuSPARSELt harnesses Sparse Tensor Cores to accelerate general matrix multiplications. Version 0.6. adds support for the NVIDIA Hopper architecture.
jwitsoe
1
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| cuSPARSELt v0.1.0 Now Available: Arm and Windows Support | 0 | 448 | April 23, 2021 | |
| Exploiting NVIDIA Ampere Structured Sparsity with cuSPARSELt | 10 | 1455 | March 14, 2022 | |
| Accelerating ReLu and GeLu Activation Functions, and Batched Sparse GEMM in cuSPARSELt v0.2.0 | 0 | 472 | November 15, 2021 | |
| Just Released: cuSPARSELt v0.3 | 0 | 291 | June 21, 2022 | |
| cuSPARSE Matrix-Vector multiplication in half precision runs slower compared to single precision | 6 | 907 | January 26, 2021 | |
| Cusparselt can't choose the best kernel(sometime)? | 4 | 467 | January 25, 2024 | |
| cusparseLtMatmul is slower than cublasGemmEx | 0 | 666 | April 21, 2023 | |
| Just Released: cuSPARSELt v0.3 | 0 | 285 | June 21, 2022 | |
| Sparse Matrix-Vector Multiplication on CUDA | 79 | 314754 | November 22, 2010 | |
| How to confirm Tensor Core is working or not in CuSPARSE | 3 | 1042 | April 28, 2023 |