Zhiyi Hu, Siyuan Shen, Tommaso Bonato, Sylvain Jeaugey, Cedell Alexander, Eric Spada, James Dinan, Jeff Hammond, Torsten Hoefler:
Demystifying NCCL: An In-depth Analysis of GPU Communication Protocols and Algorithms
(In Proceedings of the 32nd Annual Symposium on High-Performance Interconnects (HOTI'25), presented in Virtual Event, IEEE Press, Aug. 2025)
Abstract
The NVIDIA Collective Communication Library (NCCL) is a critical software layer enabling high-performance collectives on large-scale GPU clusters. Despite being open source with a documented API, its internal design remains largely opaque. The orchestration of communication channels, selection of protocols, and handling of memory movement across devices and nodes are not well understood, making it difficult to analyze performance or identify bottlenecks. This paper presents a comprehensive analysis of NCCL, focusing on its communication protocol variants (Simple, LL, and LL128), mechanisms governing intra-node and inter-node data movement, and ringand tree-based collective communication algorithms. The insights obtained from this study serve as the foundation for ATLAHS, an application-trace-driven network simulation toolchain capable of accurately reproducing NCCL communication patterns in largescale AI training workloads. By demystifying NCCL’s internal architecture, this work provides guidance for system researchers and performance engineers working to optimize or simulate collective communication at scale
Documents
download article: download slides:
Recorded talk (best effort)
BibTeX
@article{hu2025demystifying, author={Zhiyi Hu and Siyuan Shen and Tommaso Bonato and Sylvain Jeaugey and Cedell Alexander and Eric Spada and James Dinan and Jeff Hammond and Torsten Hoefler}, title={{Demystifying NCCL: An In-depth Analysis of GPU Communication Protocols and Algorithms}}, year={2025}, month={Aug.}, booktitle={Proceedings of the 32nd Annual Symposium on High-Performance Interconnects (HOTI'25)}, location={Virtual Event}, publisher={IEEE Press}, source={http://www.unixer.de/~htor/publications/}, }
