Accelerating Intra-Node GPU-to-GPU Communication Through Multi-Path Transfers with CUDA Graphs
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910168335253504 |
|---|---|
| author | Sojoodi, Amirhossein Temucin, Yiltan Hassan Baratisedeh, Amirreza Sharifian, Hamed Afsahi, Ahmad |
| author_facet | Sojoodi, Amirhossein Temucin, Yiltan Hassan Baratisedeh, Amirreza Sharifian, Hamed Afsahi, Ahmad |
| contents | Effective intra-node GPU communication is essential for optimizing performance in MPI-based HPC applications, especially when leveraging multiple communication paths. In this study, we propose a novel approach that integrates CUDA Graphs into the UCX framework to enhance intra-node multi-path point-to-point GPU communication. By concurrently leveraging multiple paths, including NVLink and PCIe through the host, and optimizing communication workflows using CUDA Graph, we achieve significant reductions in communication overhead and improve execution efficiency. To the best of our knowledge, our proposed approach is the first to seamlessly integrate CUDA Graphs into UCX. Through extensive experiments on a four-GPU node, our proposed CUDA Graph-based multi-path communication approach achieves up to a 2.95x bandwidth improvement, compared to the single-path UCX (UCT::CUDA-IPC), in GPU-to-GPU OMB bandwidth test when utilizing the host path and two other GPU paths, at message sizes up to 512MB. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_22228 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Accelerating Intra-Node GPU-to-GPU Communication Through Multi-Path Transfers with CUDA Graphs Sojoodi, Amirhossein Temucin, Yiltan Hassan Baratisedeh, Amirreza Sharifian, Hamed Afsahi, Ahmad Distributed, Parallel, and Cluster Computing Effective intra-node GPU communication is essential for optimizing performance in MPI-based HPC applications, especially when leveraging multiple communication paths. In this study, we propose a novel approach that integrates CUDA Graphs into the UCX framework to enhance intra-node multi-path point-to-point GPU communication. By concurrently leveraging multiple paths, including NVLink and PCIe through the host, and optimizing communication workflows using CUDA Graph, we achieve significant reductions in communication overhead and improve execution efficiency. To the best of our knowledge, our proposed approach is the first to seamlessly integrate CUDA Graphs into UCX. Through extensive experiments on a four-GPU node, our proposed CUDA Graph-based multi-path communication approach achieves up to a 2.95x bandwidth improvement, compared to the single-path UCX (UCT::CUDA-IPC), in GPU-to-GPU OMB bandwidth test when utilizing the host path and two other GPU paths, at message sizes up to 512MB. |
| title | Accelerating Intra-Node GPU-to-GPU Communication Through Multi-Path Transfers with CUDA Graphs |
| topic | Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2604.22228 |