Index | index by Group | index by Distribution | index by Vendor | index by creation date | index by Name | Mirrors | Help | Search |
Name: rccl | Distribution: Fedora Project |
Version: 6.3.0 | Vendor: Fedora Project |
Release: 3.fc42 | Build date: Sat Jan 18 15:52:20 2025 |
Group: Unspecified | Build host: buildvm-x86-07.iad2.fedoraproject.org |
Size: 408606894 | Source RPM: rccl-6.3.0-3.fc42.src.rpm |
Packager: Fedora Project | |
Url: https://github.com/ROCm/rccl | |
Summary: ROCm Communication Collectives Library |
RCCL (pronounced "Rickle") is a stand-alone library of standard collective communication routines for GPUs, implementing all-reduce, all-gather, reduce, broadcast, reduce-scatter, gather, scatter, and all-to-all. There is also initial support for direct GPU-to-GPU send and receive operations. It has been optimized to achieve high bandwidth on platforms using PCIe, xGMI as well as networking using InfiniBand Verbs or TCP/IP sockets. RCCL supports an arbitrary number of GPUs installed in a single node or multiple nodes, and can be used in either single- or multi-process (e.g., MPI) applications. The collective operations are implemented using ring and tree algorithms and have been optimized for throughput and latency. For best performance, small operations can be either batched into larger operations or aggregated through the API.
BSD-3-Clause AND MIT AND Apache-2.0
* Sat Jan 18 2025 Tom Rix <Tom.Rix@amd.com> - 6.3.0-3 - Add gfx1200,gfx1201 * Fri Dec 27 2024 Tom Rix <Tom.Rix@amd.com> - 6.3.0-2 - Add --with export - Remove unneeded requires rocm-rpm-macros-modules * Tue Dec 10 2024 Tom Rix <Tom.Rix@amd.com> - 6.3.0-1 - Update to 6.3
/usr/lib/.build-id /usr/lib/.build-id/9e /usr/lib/.build-id/9e/b9bb06674cc429b8ae162e420512e5a10e4897 /usr/lib64/librccl.so.1 /usr/lib64/librccl.so.1.0 /usr/share/licenses/rccl /usr/share/licenses/rccl/LICENSE.txt
Generated by rpm2html 1.8.1
Fabrice Bellet, Thu Jan 23 23:45:20 2025