mlsanityData to 14 Oct 2026

GPU-centric Communication Schemes for HPC and ML Applications

arXiv 2503.24230 ↗First posted 31 Mar 2025cs.DC cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Compute nodes on modern heterogeneous supercomputing systems comprise CPUs, GPUs, and high-speed network interconnects (NICs). Parallelization is identified as a technique for effectively utilizing these systems to execute scalable simulation and deep learning workloads. The resulting inter-process communication from the distributed execution of these parallel workloads is one of the key factors contributing to its performance bottleneck. Most programming models and runtime systems enabling the communication requirements on these systems support GPU-aware communication schemes that move the GPU-attached communication buffers in the application directly from the GPU to the NIC without staging through the host memory. A CPU thread is required to orchestrate the communication operations even with support for such GPU-awareness. This survey discusses various available GPU-centric communication schemes that move the control path of the communication operations from the CPU to the GPU. This work presents the need for the new communication schemes, various GPU and NIC capabilities required to implement the schemes, and the potential use-cases addressed. Based on these discussions, challenges involved in supporting the exhibited GPU-centric communication schemes are discussed.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 20 of its 95 references that are in the map. First posted on arXiv 31 Mar 2025.

Builds on

20 of 95 references.

References · 20

  1. 2018Demystifying Parallel and Distributed Deep Learning: An In-Depth Concurrency Analysiscited by 156
  2. 2013Deep Learning of Representations: Looking Forwardcited by 84
  3. 2019Scaling Distributed Machine Learning with In-Network Aggregationcited by 65
  4. 2021TACCL: Guiding Collective Algorithm Synthesis using Communication Sketchescited by 38
  5. 2018CosmoFlow: Using Deep Learning to Learn the Universe at Scalecited by 14
  6. 2020An In-Depth Analysis of the Slingshot Interconnectcited by 13
  7. 2024Understanding Data Movement in Tightly Coupled Heterogeneous Systems: A Case Study with the Grace Hopper Superchipcited by 8
  8. 2021A Survey and Empirical Evaluation of Parallel Deep Learning Frameworkscited by 7
  9. 2024Demystifying the Communication Characteristics for Distributed Transformer Modelscited by 7
  10. 2023Distributed Matrix-Based Sampling for Graph Neural Network Trainingcited by 6
  11. 2024Exploring GPU-to-GPU Communication: Insights into Supercomputer Interconnectscited by 6
  12. 2022MPIX Stream: An Explicit Solution to Hybrid MPI+X Programmingcited by 2
8 more
  1. 2024The Landscape of GPU-Centric Communicationcited by 2
  2. 2024Intel(R) SHMEM: GPU-initiated OpenSHMEM using SYCLcited by 2
  3. 2024Understanding Data Movement in AMD Multi-GPU Systems with Infinity Fabriccited by 2
  4. 2018Extreme Scale De Novo Metagenome Assemblycited by 1
  5. 2022HipBone: A performance-portable GPU-accelerated C++ version of the NekBone benchmarkcited by 1
  6. 2022Exploring GPU Stream-Aware Message Passing using Triggered Operationscited by 1
  7. 2023Exploring Fully Offloaded GPU Stream-Aware Message Passingcited by 1
  8. 2024Understanding GPU Triggering APIs for MPI+X Communicationcited by 1

Built on by

Not cited yet.