Saved in:
Bibliographic Details
Main Authors: Oliveira-Filho, Antônio, Silva-de-Souza, Wellington, Sakuyama, Carlos Alberto Valderrama, Xavier-de-Souza, Samuel
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.17734
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911222266331136
author Oliveira-Filho, Antônio
Silva-de-Souza, Wellington
Sakuyama, Carlos Alberto Valderrama
Xavier-de-Souza, Samuel
author_facet Oliveira-Filho, Antônio
Silva-de-Souza, Wellington
Sakuyama, Carlos Alberto Valderrama
Xavier-de-Souza, Samuel
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2502_17734
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Phoeni6: a Systematic Approach for Evaluating the Energy Consumption of Neural Networks
Oliveira-Filho, Antônio
Silva-de-Souza, Wellington
Sakuyama, Carlos Alberto Valderrama
Xavier-de-Souza, Samuel
Machine Learning
Software Engineering
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.
title Phoeni6: a Systematic Approach for Evaluating the Energy Consumption of Neural Networks
topic Machine Learning
Software Engineering
url https://arxiv.org/abs/2502.17734