We at the ML.ENERGY Initiative view energy and power that require careful measurement, understanding, allocation, and optimization across the computing system. Energy is the new first-class computing resource, and we need to build our tools, mental models, and systems around that.

Since 2021, we've built widely adopted open-source projects backed by strong computer systems research. For instance, Zeus (https://ml.energy/zeus) is part of the PyTorch Ecosystem and adopted widely by developers and researchers. We've also continued to shared knowledge with the community via our tech blog (https://ml.energy/blog) and tutorials (NeurIPS 2025; https://ml.energy/tutorials/neurips25).

Subscribe to our newsletter to stay updated: https://buttondown.com/ml-energy

ML.ENERGY

Org chart

Teams

Offices