Automated PMC-based Power Modeling Methodology for Modern Mobile GPUs

TL;DR

This paper introduces APGPM, a novel PMC-based power modeling methodology for mobile GPUs that automatically selects optimal counters, significantly improving accuracy over prior models while using fewer PMCs.

Contribution

The paper presents APGPM, an automatic PMC selection method for mobile GPU power modeling, addressing multicollinearity and enhancing accuracy over existing utilization-based models.

Findings

APGPM reduces modeling error by up to 2.66x compared to prior models.

APGPM uses only 4.66% to 20.41% of available PMCs.

Evaluation on two mobile GPUs demonstrates significant accuracy improvements.

Abstract

The rise of machine learning workload on smartphones has propelled GPUs into one of the most power-hungry components of modern smartphones and elevates the need for optimizing the GPU power draw by mobile apps. Optimizing the power consumption of mobile GPUs in turn requires accurate estimation of their power draw during app execution. In this paper, we observe that the prior-art, utilization-frequency based GPU models cannot capture the diverse micro-architectural usage of modern mobile GPUs.We show that these models suffer poor modeling accuracy under diverse GPU workload, and study whether performance monitoring counter (PMC)-based models recently proposed for desktop/server GPUs can be applied to accurately model mobile GPU power. Our study shows that the PMCs that come with dominating mobile GPUs used in modern smartphones are sufficient to model mobile GPU power, but exhibit multicollinearity if used altogether. We present APGPM, the mobile GPU power modeling methodology that automatically selects an optimal set of PMCs that maximizes the GPU power model accuracy. Evaluation on two representative mobile GPUs shows that APGPM-generated GPU power models reduce the MAPE modeling error of prior-art by 1.95x to 2.66x (i.e., by 11.3% to 15.4%) while using only 4.66% to 20.41% of the total number of available PMCs.