Google to Power Down AI Data Centers to Help Strained Grids

Google
Google's headquarters, the Googleplex.

Key Points

  • Google made deals with two U.S. utilities to reduce power use at its AI data centers.
  • The goal is to help ease the strain on the power grid during peak demand. Massive electricity demand from AI is outstripping supply in some areas.
  • This is the first time Google has a formal “demand-response” agreement for its AI workloads.
  • These types of power-sharing deals may become more common as electricity supplies tighten.

Google has struck a deal with two U.S. electric utilities to scale back power consumption at its massive AI data centers during times of high demand. The company announced the agreements on Monday as a way to help ease the growing strain that energy-hungry AI is putting on the nation’s power grid.

Utilities across the country are struggling to meet the increasing electricity demands from Big Tech’s data centers. In some regions, the power demand now outstrips the available supply, raising fears of blackouts and higher energy bills for homes and businesses. This power crunch has also become a major hurdle for the tech industry’s ambitious AI expansion plans.

Google’s new agreements with Indiana Michigan Power and the Tennessee Power Authority are the first of their kind for the company’s AI operations. Under these “demand-response” programs, Google will temporarily reduce power to its machine learning workloads when called upon by the utilities. This frees up electricity for the rest of the grid during peak usage.

In a blog post, Google said this approach helps data centers connect to the grid more quickly and reduces the need to build new power plants.
While demand-response programs are common in other energy-intensive industries like manufacturing and crypto mining, they are a new development for AI data centers.

Businesses that participate typically receive payments or lower electricity bills in return. However, the specific details of Google’s deals were not released. As the U.S. electricity supply tightens, these types of power-sharing arrangements may become a much more common solution.

EDITORIAL TEAM
EDITORIAL TEAM
Al Mahmud Al Mamun leads the TechGolly editorial team. He served as Editor-in-Chief of a world-leading professional research Magazine. Rasel Hossain is supporting as Managing Editor. Our team is intercorporate with technologists, researchers, and technology writers. We have substantial expertise in Information Technology (IT), Artificial Intelligence (AI), and Embedded Technology.
ADVERTISEMENT
3rd party Ad. Not an offer or recommendation by atvite.com.
Read More