Practical and Scalable RDMA Connection Sharing for HPC Workload
Yuejie Wang, Tuo Fang, Biyu Peng, Yang Cheng, Xin Sun, Chengchao Xu, Yuchen Tang, Yuxin Ren, Ning Jia, Xinwei Hu, Yunfei Du, Guyue (Grace) Liu
摘要
RDMA is a fundamental communication infrastructure in high-performance computing (HPC). However, as the number of RDMA connections increases, system performance rapidly declines and memory consumption increases sharply. Previous research demonstrates that sharing RDMA connections among processes is necessary and effective to address the scalability problem. Unfortunately, previous work shares connections in software, thus incurring substantial overhead to each packet operation, and fails to comprehensively explore control policies to achieve superior sharing decisions.
问问这篇 Paper
问问你的智能体。
Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。
相关 Paper
- Birds of a Feather Flock Together: Scaling RDMA RPCs with FlockSumit Kumar Monga, Sanidhya Kashyap, Changwoo MinSOSP 2021 · 被引用 34 次
- Scalable RDMA Transport with Efficient Connection SharingJian Tang, Xiaoliang Wang, Huichen DaiINFOCOM 2023 · 被引用 13 次
- Maximizing the Benefit of RDMA at End HostsXiaoliang Wang, Hexiang Song, Cam-Tu Nguyen, Dongxu Cheng 等INFOCOM 2021 · 被引用 8 次
- Network Load Balancing with In-network Reordering Support for RDMACha Hwan Song, Xin Zhe Khooi, Raj Joshi, Inho Choi 等SIGCOMM 2023 · 被引用 110 次
- Scalable RDMA-accelerated Distributed Locks with Shared Stream AbstractionMiao Cai, Junru Shen, Xiaojian Liao, Rong Gu 等EuroSys 2026
