Predict Before Replay: Joint FEC and Flight Control for Reliable Scale-Up Links

📅 2026-08-18
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
本文提出PREFACE方法,通过预FEC控制器解决高速链路中数据重传效率低的问题,提高了吞吐量并降低了延迟。
📝 Abstract
Scale-up accelerator fabrics send latency-sensitive flits over serial links at hundreds of gigabits per second. Their reliability pipeline first relies on FEC, then detects residual failures and replays unacknowledged data. At these line rates, delayed feedback lets later flits enter the replay window before a residual failure is reported, so standard replay can amplify one corrupted flit into a suffix retransmission. This paper presents PREFACE, a pre-FEC controller for temporally correlated burst errors. A two-state Bayesian filter converts corrected-symbol observations into a next-flit burst posterior and jointly selects FEC strength with an outstanding-flit cap. We implement PREFACE in ns-3 with publicly verifiable UALink 200G 1.0 replay semantics. PREFACE improves goodput by 10.52%, lowers P99 latency by 50.75%, cuts replay by 47.52%, and improves modeled ring AllReduce by 13.1--27.0%.
Problem

Research questions and friction points this paper is trying to address.

FEC
replay
latency
burst errors
serial links
Innovation

Methods, ideas, or system contributions that make the work stand out.

pre-FEC controller
Bayesian filter
burst error
FEC strength
replay reduction
F
Fan Yang
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China
J
Jiaqi Liu
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China; Hangzhou Institute for Advanced Study, UCAS, Hangzhou, China
Tao Jiang
Tao Jiang
Beijing Institute of Nanoenergy and Nanosystems
Nanogeneratorblue energyenergy harvestingself-assembly
Z
Zhan Wang
Institute of Computing Technology, Chinese Academy of Sciences, Beijing, China