Phoeni6: a Systematic Approach for Evaluating the Energy Consumption of Neural Networks

TL;DR

Phoeni6 is a systematic framework for evaluating neural network energy consumption, promoting fair, reproducible, and transparent assessments across different models, data, and configurations.

Contribution

It introduces a comprehensive, automated methodology for energy evaluation of neural networks, ensuring portability, transparency, and reproducibility in energy consumption assessments.

Findings

MobileNet is up to 6.25% more energy-efficient than AlexNet.

BMP images reduce energy consumption by up to 30% compared to PNG.

Phoeni6 facilitates sustainable AI practices through standardized energy evaluation.

Abstract

This paper presents Phoeni6, a systematic approach for assessing the energy consumption of neural networks while upholding the principles of fair comparison and reproducibility. Phoeni6 offers a comprehensive solution for managing energy-related data and configurations, ensuring portability, transparency, and coordination during evaluations. The methodology automates energy evaluations through containerized tools, robust database management, and versatile data models. In the first case study, the energy consumption of AlexNet and MobileNet was compared using raw and resized images. Results showed that MobileNet is up to 6.25% more energy-efficient for raw images and 2.32% for resized datasets, while maintaining competitive accuracy levels. In the second study, the impact of image file formats on energy consumption was evaluated. BMP images reduced energy usage by up to 30% compared to PNG, highlighting the influence of file formats on energy efficiency. These findings emphasize the importance of Phoeni6 in optimizing energy consumption for diverse neural network applications and establishing sustainable artificial intelligence practices.