Google Faces Decade-Long Waits to Power New AI Data Centers

google
Google's headquarters, the Googleplex. [TechGolly]

Key Points

  • Google is facing wait times of over a decade to connect new data centers to the U.S. power grid.
  • These “transmission barriers” are the number one challenge for powering the AI boom.
  • To get around the delays, Google is exploring “co-location,” or building data centers next to power plants.
  • Co-location is a controversial solution that raises cost and reliability concerns.

Google is encountering a major roadblock in its quest to power the artificial intelligence boom: the U.S. electrical grid. According to a Google energy executive, the wait times to connect a new data center to the transmission system have surged to more than a decade in some parts of the country.

“Transmission barriers are the number one challenge we’re seeing on the grid,” Marsden Hanna, Google’s Global Head of Sustainability and Climate Policy, said on Wednesday. “We had one utility who told us 12 years to study the interconnection timeline, which is sort of wild, but that’s what we’re seeing.”

This is a significant challenge for Google and other tech giants that are racing to build energy-intensive data centers to train and run AI models. The country’s power system is simply not keeping up with the explosive demand.

To address these long delays, Hanna said Google is now exploring a strategy called “co-location.” This would involve building data centers adjacent to power plants, allowing them to bypass the transmission system and its long wait times.

However, co-location is a complex and controversial solution. It raises questions about who should pay for the costs of these arrangements and what it means for the rest of the grid when a large amount of power is diverted to a single customer.

Federal and regional regulators are now investigating the issue and developing guidelines for this new approach. In the meantime, Hanna said the country needs to expedite the permitting process for new transmission lines and adopt new technologies to increase power output from the existing system. Without these changes, the AI revolution could be slowed down by something as simple as a lack of power.

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.
Read More