Modeling and Chasing the Energy Efficiency Sweet Spots in Modern GPUs

Jul 16, 2026

Upload by Anda

Energy consumption limits high-performance computing on heterogeneous CPU–GPU systems, making energy-to-solution highly sensitive to hardware configuration. This work analyzes energy-efficiency regimes under realistic molecular dynamics workloads (GROMACS, AMBER) and a stress-test benchmark (FIRESTARTER) across A40, A100, H100, and H200 GPUs and an Intel Ice Lake CPU while varying DVFS frequency scaling and power caps. It reveals workload- and architecture-dependent transitions between efficient and inefficient regimes driven by nonlinear power-frequency scaling, and introduces an interpretable model decomposing GPU power into baseline, linear, and nonlinear terms to identify a transition frequency. Frequency scaling outperforms power capping, especially when workloads run far below thermal design power.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

Introduction and related work

Contributions

Overview

Workloads, testbed and experimental setup

arXiv :2607 .008 19v 1 [ cs .DC] 1 Jul 2026 Modeling and Chasing the Energy-Efficiency Sweet Spots in Modern GPUs Ayesha Afzal 1[0000−0001−5061−0438], Markus Manfred Li2[0009−0004−0908−3460], and Michael Panzlaff1[0009−0009−2993−7001] 1 Erlangen National High Performance Computing Center (NHR@FAU) 2 Department of Computer Science, FAU Erlangen-Nürnberg, Germany Abstract. Energy consumption is a key limitation in high-performance computing on heterogeneous CPU–GPU systems. This work studies how hardware configuration affects energy-to-solution under realistic workloads. We study energy efficiency regimes using molecular dynamics benchmarks (GROMACS and AMBER) and a stress-test benchmark (FIRESTARTER) on systems with A40, A100, H100, and H200 GPUsand Intel Ice Lake CPU, varying frequency scaling and power cap. We show that energy-to-solution exhibits workload-and architecture-dependent transitions between efficient and inefficient regimes, driven by nonlinear GPU power-frequency scaling. We introduce an interpretable analytical model that decomposes GPU power into linear and nonlinear components, identifying a workload-and architecture-dependent transition frequency beyond which efficiency degrades. The model fits empirical data with low error and highlights the role of baseline power, nonlinear power behavior, and transition frequency as the dominant parameters governing energy efficiency. Power capping is generally less effective for efficiency tuning than frequency reduction, especially for workloads that operate far from thermal design power. Overall, energy-efficient HPC execution is a configuration-dependent problem with identifiable regime shifts, and we provide model-driven guidance for selecting operating points. Keywords: energy efficiency · molecular dynamics · frequency scaling · power cap · power modeling · stress test FIRESTARTER · GROMACS · AMBER 1 Introduction and related work Energy efficiency has emerged as a primary limiting factor in high-performance computing (HPC), increasingly constraining system scalability under fixed power budgets rather than peak computational capability. This challenge is particularly pronounced in heterogeneous CPU–GPU systems, where multiple hardware control parameters jointly determine performance and energy to solution. A central difficulty arises from the high-dimensional configuration space exposed by modern HPC systems, including GPU dynamic voltage and frequency scaling (DVFS) [12], power caps [10], and empirical power modeling [11,15,14,13] . These strategies exhibit nonlinear GPU power scaling due to voltage–frequency coupling and workload dependence, yet most studies rely on synthetic kernels, limit- 2 A. Afzal et al. ing their applicability to real-world applications. Few existing approaches implicitly assume smooth and monotonic relationships between performance, power, and energy efficiency. However, such assumptions frequently break down for scientific workloads with heterogeneous compute and memory behavior. Empirical or machine-learning-based energy models exist but often lack interpretability and hardware-level insight, creating a gap between predictive accuracy and physical understanding for realistic HPC workloads. In this work, we focus on molecular dynamics (MD) applications, represented by GROMACS [1,6] and AMBER [7], complemented by the synthetic stress-test benchmark FIRESTARTER [8] . These workloads are executed on multiple GPU architectures (A40, A100, H100, H200) and an Intel Ice Lake CPU platform. Collectively, they span compute-bound, memory-bound, and thermally constrained regimes, enabling a unified analysis of energy behavior across diverse operating conditions. Contributions This paper makes the following contributions: – We systematically evaluate energy-to-solution using molecular dynamics workloads (GROMACS, AMBER) and a synthetic stress-test (FIRESTARTER), covering GPU frequency scaling and power cap configurations. – We in

FAQs

What problem does the paper address in modern heterogeneous CPU–GPU HPC systems?

The paper studies how hardware configuration affects energy-to-solution, since energy consumption increasingly constrains scalability under fixed power budgets.

Which benchmarks and hardware platforms are used to evaluate energy efficiency regimes?

It uses molecular dynamics workloads (GROMACS, AMBER) and the stress-test benchmark FIRESTARTER on GPUs A40, A100, H100, H200 and an Intel Ice Lake CPU, while varying frequency scaling and power cap settings.

How does the paper explain the observed drops in efficiency?

It shows that energy-to-solution can transition between efficient and inefficient regimes, driven by nonlinear GPU power-frequency scaling; an analytic model identifies a workload- and architecture-dependent transition frequency beyond which efficiency degrades.

Jul 16, 2026

Upload by Anda