arXiv · 2611.17206
PACER: GPU Excursion Management for ML Workloads
Abstract
Variability is an inherent characteristic of modern GPU systems, arising from a combination of process variation, dynamic environmental conditions, and power management decisions. Prior work has characterized inter-GPU variability and found consistent performance differences across GPUs. However, we observe that modern workloads, particularly large language models (LLMs), exhibit significant intra-GPU variability. Their short-running kernels interact poorly with the coarse granularity of dynamic power management mechanisms, leading to unpredictable performance. A key source of this intra-GPU transient variability is power excursions, also referred to as electrical design power peaks (EDP$_{p}$), where power temporarily exceeds the GPU's rated power limits. While power management mechanisms are designed to ensure reliable operation despite the presence of EDP$_{p}$ sustained power peaks cause throttling that detrimentally impacts overall application performance. To the best of our knowledge, our work is the first to characterize the magnitude and frequency of EDP$_{p}$, showing that frequent, significant power overshoots are especially common in ML workloads, which combine high power draw, short kernels, and rapidly changing compute demands. To address this, we design and leverage hardware and firmware mechanisms to provide consistent power management behavior that reduces intra-GPU variability due to EDP$_{p}$ and improves overall performance. We evaluate our techniques across a range of popular workloads. Our approach improves performance by 3-12% over state-of-the-art GPU power management and comes within 3% of an oracular solution.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Rutwik Jain, Nathaniel Morris, Yiwei Jiang, Shidhartha Das, Gary Kao, Indrani Paul, Shivaram Venkataraman, Srilatha Manne, Matthew D. Sinclair. 2026-11-12. PACER: GPU Excursion Management for ML Workloads. https://arxiv.science/abs/2611.17206
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite arXiv Science by Indiciae: BibTeX · RIS.