AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Updated RCCL introduction at docs/source/library.rst (#680)

Browse Source 
* Updated RCCL introduction at docs/source/library.rst

* space after period

---------

Co-authored-by: Saad Rahim <44449863+saadrahim@users.noreply.github.com>
This commit is contained in:
Pedram Alizadeh
2023-02-10 17:20:54 -05:00
committed by GitHub
parent df46645ff8
commit e05560ea82
1 changed files with 2 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/source/library.rst
+2 -3
View File
@@ -10,7 +10,6 @@ RCCL
Introduction
------------
The ROCm Collective Communication Library (RCCL) is a stand-alone library which provides multi-GPU and multi-node collective communication primitives optimized for AMD GPUs and can be easily integrated into applications.
RCCL (pronounced “Rickel”) implements routines such as all-reduce, all-gather, reduce, broadcast, reduce-scatter, gather, scatter, all-to-allv, and all-to-all as well as direct point-to-point (GPU-to-GPU) send and receive operations. The provided collective communication routines are implemented using Ring and Tree algorithms and are optimized to achieve high bandwidth and low latency over PCIe and xGMI high-speed interconnects for intra-node communication and over InfiniBand, RoCE, and TCP/IP for inter-node communication.
RCCL supports an arbitrary number of GPUs installed in a single-node or multi-node platform and can be used in either single- or multi-process (e.g., MPI) applications.
The ROCm Collective Communication Library (RCCL) is a stand-alone library which provides multi-GPU and multi-node collective communication primitives optimized for AMD GPUs. RCCL (pronounced “Rickel”) implements routines such as all-reduce, all-gather, reduce, broadcast, reduce-scatter, gather, scatter, all-to-allv, and all-to-all as well as direct point-to-point (GPU-to-GPU) send and receive operations. The provided collective communication routines are implemented using Ring and Tree algorithms. They are optimized to achieve high bandwidth and low latency by leveraging topology awareness, high-speed interconnects, RDMA based collectives. RCCL utilizes PCIe and xGMI high-speed interconnects for intra-node communication as well as InfiniBand, RoCE, and TCP/IP for inter-node communication. 
RCCL supports an arbitrary number of GPUs installed in a single-node or multi-node platform. It can be easily integrated into either single- or multi-process (e.g., MPI) applications.