# HiCCL: A Hierarchical Collective Communication Library

arXiv: https://arxiv.org/abs/2408.05962 · HTML: https://mlsanity.com/paper/2408.05962/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Aug 2024.

## Built on by (most cited first, 3 of 3)

- [Bine Trees: Enhancing Collective Operations by Optimizing Communication Locality](https://mlsanity.com/paper/2508.17311/) (2025)
- [HetCCL: Accelerating LLM Training with Heterogeneous GPUs](https://mlsanity.com/paper/2601.22585/) (2026)
- [The Big Send-off: Scalable and Performant Collectives for Deep Learning](https://mlsanity.com/paper/2504.18658/) (2025)

## Abstract (arXiv)

HiCCL (Hierarchical Collective Communication Library) addresses the growing complexity and diversity in high-performance network architectures. As GPU systems have envolved into networks of GPUs with different multilevel communication hierarchies, optimizing each collective function for a specific system has become a challenging task. Consequently, many collective libraries struggle to adapt to different hardware and software, especially across systems from different vendors. HiCCL's library design decouples the collective communication logic from network-specific optimizations through a compositional API. The communication logic is composed using multicast, reduction, and fence primitives, which are then factorized for a specified network hieararchy using only point-to-point operations within a level. Finally, striping and pipelining optimizations applied as specified for streamlining the execution. Performance evaluation of HiCCL across four different machinesunicodex2014two with Nvidia GPUs, one with AMD GPUs, and one with Intel GPUsunicodex2014demonstrates an average 17× higher throughput than the collectives of highly specialized GPU-aware MPI implementations, and competitive throughput with those of vendor-specific libraries (NCCL, RCCL, and OneCCL), while providing portability across all four machines.