The alliance, which includes Google, Nvidia, Anthropic, and major utility providers like AES and National Grid, focuses on managing power consumption during peak hours. Rather than relying on traditional methods like diesel generators, Emerald AI’s platform enables data centers to pause noncritical tasks or move compute loads to regions with available grid headroom. This real-time coordination transforms data centers into flexible assets, similar to large-scale batteries, capable of responding instantly to utility signals.
In section Startups & Technology
Google and Nvidia back Emerald AI coalition to solve data center power
A new industry coalition, the AI Energy Management Alliance, is launching to bridge the gap between massive data center electricity demand and aging power grids. By using Emerald AI’s software, tech giants and utilities aim to unlock 100 gigawatts of capacity through synchronized demand response and load shifting.

While demand response is a established practice in industrial sectors, the explosive growth of AI compute loads has made this integration critical for future infrastructure development. Goldman Sachs estimates that limiting maximum grid usage could free up 76 gigawatts of capacity, a figure the coalition hopes to exceed. Despite the technological leap, Emerald AI chief scientist Ayse Coskun notes that these efficiencies will reduce the industry’s reliance on new power plants but will not entirely eliminate the need for new generating sources. Following a $150 million Series A funding round, the startup is now positioned to deploy this orchestration software across the utility sector at scale.
Comments (0)
No comments yet. Be the first!