KEY POINTS
- AI is driving data center energy demand to historic levels.
- U.S. electricity use by data centers could reach 12 percent by 2028.
- The DOE explores grid resilience measures like advanced storage and nuclear tech.
U.S. data centers could consume as much as 12 percent of the nation’s electricity by 2028, nearly tripling their current demand.
This projection, outlined in a Department of Energy-backed report from the Lawrence Berkeley National Laboratory, highlights the transformative impact of artificial intelligence (AI) on the industry.
AI’s growing reliance on advanced GPUs and high-intensity cooling systems has accelerated energy needs. “The frontier of energy demand growth in the U.S. is clearly AI data centers,” said Avi Shultz, director of the DOE’s Industrial Efficiency and Decarbonization Office.
According to Reuters, the report estimates that AI data centers’ annual energy consumption could soar to between 74 and 132 gigawatts, equivalent to 6.7 percent to 12 percent of total U.S. electricity usage.
Grid resilience and energy efficiency take center stage
The dramatic rise in AI-driven electricity use is prompting discussions about grid resilience and efficiency.
The DOE is considering measures such as long-duration battery storage and advanced technologies like small nuclear reactors to support this growth.
Report findings also stress the need for real-time data monitoring. Lead researcher Arman Shehabi suggests publishing updates annually to track trends and identify energy-saving opportunities. “Understanding what’s driving this growth helps pinpoint efficiency strategies,” Shehabi noted.
Data center innovations and new power challenges
Between 2017 and 2023, the deployment of GPU-accelerated servers more than doubled the energy demands of data centers. AI servers now account for a significant portion of this surge, requiring both advanced chips and robust cooling systems.
The report highlights that new AI data centers are being constructed with power capacities reaching one gigawatt, enough to power a city like Philadelphia.
As U.S. power demand hits record levels, AI’s impact on energy consumption and grid sustainability is becoming a central focus for policymakers.
