Characterizing Software Aging in GPU-Based LLM Serving Systems
Not on the map. Why?
Abstract (arXiv)
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.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 7 of its 24 references that are in the map. First posted on arXiv 10 Jun 2026.
Builds on
7 of 24 references.
References · 7
- 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
- 2024vAttention: Dynamic Memory Management for Serving LLMs without PagedAttentioncited by 98
- 2025Jenga: Effective Memory Management for Serving LLM with Heterogeneitycited by 25
- 2024SYMPHONY: Improving Memory Management for LLM Inference Workloadscited by 5
- 2025Investigating Software Aging in LLM-Generated Software Systemscited by 2
- 2021Software Micro-Rejuvenation for Android Mobile Systemscited by 1
- 2025Adaptive Detection of Software Aging under Workload Shiftcited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Measurement-Driven Diagnosis and Mitigation of Host-CPU Co-location Interference in Single-GPU LLM Serving on a Multi-GPU Server ↗
- 2025Demystifying Cost-Efficiency in LLM Serving over Heterogeneous GPUs
- 2026LLMServingSim 2.0: A Unified Simulator for Heterogeneous and Disaggregated LLM Serving Infrastructure
- 2026Lightweight Scheduling Policies for Continuously Batched LLM Serving Across GPU Architectures
- 2024KunServe: Parameter-centric Memory Management for Efficient Memory Overloading Handling in LLM Serving
- 2026SiliconBench: Speed, Memory, and Fidelity for LLM Serving on Unified-Memory Desktops
- 2024ConServe: Fine-Grained GPU Harvesting for LLM Online and Offline Co-Serving
- 2024Saving GPU Hours in LLM Inference System Development and Online Workloads with Simulation and DBMS-Inspired Cache Replacement Policies