paper-with-me

Papers

Taming the Chaos: Coordinated Autoscaling for Heterogeneous and Disaggregated LLM Inference

2025-08-27 · Rongzhi Li, Ruogu Du, Zefang Chu, Sida Zhao, Chunlei Han, Zuocheng Shi, Yiwen Shao, Huanle Han, Long Huang, Zherui Liu, Shufan Liu arxiv

Serving Large Language Models (LLMs) is a GPU-intensive task where traditional autoscalers fall short, particularly for modern Prefill-Decode (P/D) disaggregated architectures. This architectural shift, while powerful, introduces significant operational challenges, including inefficient use of heterogeneous hardware, network bottlenecks, and critical imbalances between prefill and decode stages. We introduce HeteroScale, a coordinated autoscaling framework that addresses the core challenges of P/D disaggregated serving. HeteroScale combines a topology-aware scheduler that adapts to heterogeneous hardware and network constraints with a novel metric-driven policy derived from the first large-scale empirical study of autoscaling signals in production. By leveraging a single, robust metric to jointly scale prefill and decode pools, HeteroScale maintains architectural balance while ensuring efficient, adaptive resource management. Deployed in a massive production environment on tens of thousands of GPUs, HeteroScale has proven its effectiveness, increasing average GPU utilization by a significant 26.6 percentage points and saving hundreds of thousands of GPU-hours daily, all while upholding stringent service level objectives.

📄 PDF Abstract BibTeX arXiv:2508.19559

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Learning in Chaos: Efficient Autoscaling and Self-healing for Distributed Training at the Edge

2025-05-19 · Wenjiao Feng, Rongxing Xiao, Zonghang Li, Hongfang Yu 외

Frequent node and link changes in edge AI clusters disrupt distributed training, while traditional checkpoint-based recovery and cloud-centric autoscaling are too slow for scale-out and ill-suited to chaotic and self-gov…

Scheduling

Disaggregation of Remotely Sensed Soil Moisture in Heterogeneous Landscapes using Holistic Structure based Models

2015-01-30 · Subit Chakrabarti, Jasmeet Judge, Anand Rangarajan, Sanjay Ranka

In this study, a novel machine learning algorithm is presented for disaggregation of satellite soil moisture (SM) based on self-regularized regressive models (SRRM) using high-resolution correlated information from auxil…

Clustering

FlowTrain: Flow-Based Decoupled Training for Industrial-Grade Vision-Language Models

2026-06-22 · Zhida Jiang, Zhaolong Xing, Yang Pei, Xiaolong Chen 외 arxiv

Industrial-grade distributed training of vision-language models (VLMs) remains far less efficient than that of unimodal LLMs. Existing solutions either follow a monolithic design that assigns uniform parallelism to heter…

StreamRL: Scalable, Heterogeneous, and Elastic RL for LLMs with Disaggregated Stream Generation

2025-04-22 · Yinmin Zhong, Zili Zhang, Xiaoniu Song, Hanpeng Hu 외

Reinforcement learning (RL) has become the core post-training technique for large language models (LLMs). RL for LLMs involves two stages: generation and training. The LLM first generates samples online, which are then u…

Reinforcement Learning (RL)Scheduling

Effect or Treatment Heterogeneity? Policy Evaluation with Aggregated and Disaggregated Treatments

2021-10-04 · Phillip Heiler, Michael C. Knaus

Binary treatments are often ex-post aggregates of multiple treatments or can be disaggregated into multiple treatment versions. Thus, effects can be heterogeneous due to either effect or treatment heterogeneity. We propo…