How to estimate carbon footprint when training deep learning models? A guide and review

Abstract

Machine learning and deep learning models have become essential in the recent fast development of artificial intelligence in many sectors of the society. It is now widely acknowledge that the development of these models has an environmental cost that has been analyzed in many studies. Several online and software tools have been developed to track energy consumption while training machine learning models. In this paper, we propose a comprehensive introduction and comparison of these tools for AI practitioners wishing to start estimating the environmental impact of their work. We review the specific vocabulary, the technical requirements for each tool. We compare the energy consumption estimated by each tool on two deep neural networks for image processing and on different types of servers. From these experiments, we provide some advice for better choosing the right tool and infrastructure.

Environmental Research Communications, 2023

References in corpus (1)

• Energy Usage Reports: Environmental awareness as part of algorithmic accountability

Cited by in corpus (7)

• Optical Generative Models
• Empirical Measurements of AI Training Power Demand on a GPU-Accelerated Node
• Greening AI-enabled Systems with Software Engineering: A Research Agenda for Environmentally Sustainable AI Practices
• Toward Sustainable Generative AI: A Scoping Review of Carbon Footprint and Environmental Impacts Across Training and Inference Stages
• First-of-its-kind AI model for bioacoustic detection using a lightweight associative memory Hopfield neural network
• Empirically-Calibrated H100 Node Power Models for Reducing Uncertainty in AI Training Energy Estimation
• Estimating Deep Learning energy consumption based on model architecture and training environment