NCCL (pronounced "Nickel") is a stand-alone library of standard collective communication routines, such as all-gather, reduce, broadcast, etc., that have been optimized to achieve high bandwidth over PCIe. NCCL supports an arbitrary number of GPUs installed in a single node and can be used in either single- or multi-process (e.g., MPI) applications.
| Uploaded | Mon Mar 31 00:14:04 2025 |
| md5 checksum | 041f6b9c59d9716ca66e3ba8973b29e0 |
| arch | x86_64 |
| build | cuda9.2_0 |
| depends | cudatoolkit >=9.2,<9.3.0a0, libgcc-ng >=7.3.0, libstdcxx-ng >=7.3.0 |
| license | BSD 3-Clause |
| license_family | BSD |
| md5 | 041f6b9c59d9716ca66e3ba8973b29e0 |
| name | nccl |
| platform | linux |
| sha1 | d7c19be16c8c1ff6bc3e53e557798b4d874cf130 |
| sha256 | 880264d51ad5bd5cc1793811b1553e9a2cec4afaf8520016f941401a8c57d4d4 |
| size | 1382366 |
| subdir | linux-64 |
| timestamp | 1548262779624 |
| version | 1.3.5 |