PACC: A Proactive CNP Generation Scheme for Datacenter Networks

Zhang, Jiao; Wang, Yuqing; Zhong, Xiaolong; Yu, Mingxuan; Pan, Haoyu; Zhang, Yali; Guan, Zixuan; Che, Biyao; Wan, Zirui; Pan, Tian; Huang, Tao

doi:10.1109/tnet.2024.3361771

Cited by 2 publications

(1 citation statement)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The current AI infrastructure utilizes RDMA to fulfill its low-latency, high-throughput network communication requirements. To fully leverage computational resources, AI training tasks distribute model parameters and datasets across various network nodes for distributed training [21,29,30]. Subsequently, RDMA transfers are invoked through collective communication operations such as All-to-All and All-Reduce to synchronize training results across all nodes.…”

Section: Introductionmentioning

confidence: 99%

An Optimized RDMA QP Communication Mechanism for Hyperscale AI Infrastructure

Wang,

Lin,

Zhang

et al. 2024

Preprint

Self Cite

View full text Add to dashboard Cite

The current artificial intelligence (AI) infrastructure widely employs remote direct memory access (RDMA) protocol for high-performance communication in networks, utilizing Reliable Connection (RC)-based Queue Pairs (QP) to ensure end-to-end correct and ordered data transmission. However, as the scale of AI infrastructure continues to expand, this RC-based QP communication mechanism faces deficiencies in scalability and is prone to congestion, resulting in degraded network transfer performance. In this paper, we propose an optimized RDMA QP communication mechanism to address scalability and congestion issues in hyper-scale AI infrastructure networks. Firstly, we replace RC-based QPs with Reliable Datagram (RD)-based QPs and propose a new reliable mechanism to address scalability problems, eliminating the need for repetitive QP establishment by AI processes during external communication. Additionally, to mitigate congestion caused by a single path, we implement multipath data transmission by introducing a new unordered reception method in the network software stack. Through experiments and simulation tests, the optimized RDMA QP communication in large-scale AI infrastructure exhibits exceptional scalability and significantly reduces the occurrence of congestion, resulting in an overall network performance improvement of over 15%.

show abstract

Section: Introductionmentioning

confidence: 99%