Technology

The work was done in Engineering Research Center of Key Software Technologies for Smart City Perception and Planning.Corresponding author.Corresponding author

Papers on Lattice

Total citations

Topics

h-index

Publication activitypapers/week, last 8 weeks

Research focus

Architecture Design (Transformers, SSMs, MoE) (1)Distributed Systems & Hardware (1)Inference & Quantization (1)

Frequent co-authors

Yijun Sun (1)Xudong Liao (1)Songrun Xie (1)Songru Xie (1)

Papers (1)

Mar 18, 2026

2w ago

Multi-stage Flow Scheduling for LLM Serving

LLM serving systems can boost Time-To-First-Token (TTFT) attainment by up to 2.4x simply by prioritizing network flows based on a novel approximation of Least-Laxity-First scheduling.

Yijun Sun, Xudong Liao, Songrun Xie +10

Architecture Design (Transformers, SSMs, MoE)Distributed Systems & Hardware Inference & Quantization

Search

Technology

Publication activitypapers/week, last 8 weeks

Research focus

Frequent co-authors

Papers (1)