Laser: Unlocking Layer-Level Scheduling for Efficient Multi-SLO LLM Serving

Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming 2026

View at publisher
Authors: Jianxiong Liao, Quanxing Dong, Yunkai Liang, Zhi Zhou, Xu Chen
DOI: 10.1145/3774934.3786413
Cited by: 1
References: 8

Abstract

Engaging applications with diverse SLO requirements has become indispensable for production-scale LLM serving systems. However, existing systems rely on iteration-level scheduling, which enforces inflexible, unified execution across multi-SLO workloads, significantly constraining the serving efficiency.

Fields of Study & Subjects

Optical Network Technologies Advanced Optical Network Technologies Semiconductor Lasers and Optical Devices

Publication details

Journal / source
Proceedings of the 31st ACM SIGPLAN Annual Symposium on Principles and Practice of Parallel Programming
Publisher
ACM
Pages
509–521
Publication date
28 January 2026