发现GPU大模型服务系统存在显著内存老化，且与运行时长和配置强相关。
Characterizing Software Aging in GPU-Based LLM Serving Systems
- 通过216小时多部署实测，建立针对大模型服务的软件老化研究方法。
- 所有部署均出现显著内存泄漏，泄漏速率随运行时间和配置变化。
- 提供可复现框架，适合关注系统稳定性与长期运行的研究者。
本文提出一种面向基于GPU的大模型服务系统的软件老化实证研究方法。传统老化研究多聚焦于以CPU为中心、负载相对稳定的软件系统；而大模型服务涉及Python主机与CUDA设备协同，请求开销差异可达数个数量级，且依赖快速演进的软件栈。研究在六个共置部署上开展216小时的相同压力测试，同步监控主机、设备及客户端指标，并采用考虑自相关性和多重检验的统计分析流程。结果表明，所有部署均存在显著的内存老化现象，泄漏速率与服务运行时间及部署配置密切相关。除上述发现外，本文还提供了一个可复现的研究框架，为软件老化与重启机制研究与大模型服务领域的交叉研究开辟新方向。
原文摘要 · Abstract (English)
This paper proposes an empirical methodology to study software aging in GPU-based LLM serving systems. Traditional aging studies focus on CPU-centric software with relatively regular workloads; LLM serving is different, spanning a Python host and a CUDA device, handling requests whose cost varies by orders of magnitude, and relying on rapidly evolving software stacks. We run a 216-hour campaign across six co-located deployments under identical stress conditions, monitor host, device, and client metrics in parallel, and apply a statistical pipeline that accounts for autocorrelation and multiple testing. Our results reveal statistically significant memory aging in all deployments, with leak rates strongly dependent on the serving runtime and deployment configuration. Beyond these findings, we provide a reproducible framework that opens a research direction at the intersection of the software aging and rejuvenation and LLM serving communities.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。