MatchRDMA: A Segmented and Rate-Matched Long-Haul RDMA Scheme for Geo-distributed LLM Training over OTN
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866914509415776256 |
|---|---|
| author | Dai, Jun Wang, Xiaorun Li, Xingde Yang, Zheng Fang, Kexiong Gu, Zhiqun Wang, Hongxiang Ji, Yuefeng Zhang, Jiawei |
| author_facet | Dai, Jun Wang, Xiaorun Li, Xingde Yang, Zheng Fang, Kexiong Gu, Zhiqun Wang, Hongxiang Ji, Yuefeng Zhang, Jiawei |
| contents | We propose MatchRDMA, a proactive, segmented, and rate-matched long-haul RDMA scheme for geo-distributed LLM training over OTN. By coordinating source and destination OTN rates, it improves inter-DC throughput by up to 20x compared with conventional RDMA, and reduces destination-OTN buffer occupancy by up to 62.7%. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_23932 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | MatchRDMA: A Segmented and Rate-Matched Long-Haul RDMA Scheme for Geo-distributed LLM Training over OTN Dai, Jun Wang, Xiaorun Li, Xingde Yang, Zheng Fang, Kexiong Gu, Zhiqun Wang, Hongxiang Ji, Yuefeng Zhang, Jiawei Networking and Internet Architecture We propose MatchRDMA, a proactive, segmented, and rate-matched long-haul RDMA scheme for geo-distributed LLM training over OTN. By coordinating source and destination OTN rates, it improves inter-DC throughput by up to 20x compared with conventional RDMA, and reduces destination-OTN buffer occupancy by up to 62.7%. |
| title | MatchRDMA: A Segmented and Rate-Matched Long-Haul RDMA Scheme for Geo-distributed LLM Training over OTN |
| topic | Networking and Internet Architecture |
| url | https://arxiv.org/abs/2604.23932 |