Measuring the Carbon Intensity of AI in Cloud Instances

23 February 2023

Sustainable

Research

Conducted by :

Allen Institute for AI, Carnegie Mellon University, Microsoft Research, Stanford University, UC Berkeley

#Environmental Impact #AI #Cloud Computing #Carbon Footprint #Measurement

Authors

Jesse Dodge (Allen Institute for AI)

Taylor Prewitt (Carnegie Mellon University)

Remi Tachet des Combes (Microsoft Research)

Omar Khattab (Stanford University)

Peter Henderson (Stanford University)

Neel Kant (Stanford University)

Alex Peysakhovich

Sergey Levine (UC Berkeley)

Abstract

This paper presents a methodology for accurately measuring the carbon emissions of AI workloads running in cloud environments.

The research provides detailed measurements across different cloud providers and regions, showing how carbon intensity can vary significantly based on location and time of day.

The authors also release tools and best practices for researchers and practitioners to measure and reduce the carbon footprint of their AI applications.

Sources

Page source

Document source

Notice something missing or incorrect?
Suggest changes on GitHub