University of Michigan engineers have launched new open-source software and an online leaderboard that reveal how much electricity different AI models consume. Led by Mosharaf Chowdhury, the team found that energy use can vary dramatically depending on how a model is designed and deployed. As AI’s footprint on the electric grid grows, this research gives developers new tools to cut costs and reduce environmental impact.
We use cookies to ensure you get the best experience on our website. By using this site, you accept our use of cookies.