Global data center electricity consumption is projected to surge 26 percent this year, reaching 565 terawatt-hours as AI workloads push power demand to unprecedented levels, according to new forecasts from research firm Gartner.
Tom’s Hardware reports that worldwide data center electricity consumption will climb from 447 terawatt-hours in 2025 to 565 terawatt-hours in 2026, marking a 26 percent increase driven primarily by compute-intensive AI applications. The research indicates that power availability has emerged as a critical constraint limiting AI expansion across the industry.
According to the Gartner forecast, worldwide power demand is expected to increase 27 percent to 132 gigawatts over the same period, up from 104 gigawatts in 2025. The firm projects that total consumption will surpass 1,200 terawatt-hours by 2030. The gigawatt measurement reflects peak capacity that requires construction, permitting, and grid connection, while terawatt-hours measure actual electricity drawn throughout the year. Both metrics are rising faster than utilities can expand supply infrastructure.
“Surging demand for compute-intensive AI workloads is driving unprecedented data center power growth, while AI capacity is now constrained by power availability, making data center power security the new battle ground for scaling and protecting margins in the global AI race,” said Gartner Direct Analyst Linglan Wang.
The research reveals a dramatic shift in power consumption patterns between AI-optimized and conventional server infrastructure. AI-optimized servers consumed approximately 95 terawatt-hours globally in 2025 and are projected to draw 175 terawatt-hours in 2026, representing an 84 percent increase. Gartner anticipates this figure will reach 258 terawatt-hours in 2027, marking the first time AI-optimized hardware will consume more electricity than conventional servers.
AI-optimized servers currently represent 31 percent of total data center power consumption in 2026, up from roughly 20 percent the previous year. Cooling systems also consume a growing portion of electricity, with power used for cooling forecast to climb 22.6 percent in 2026 to 195 terawatt-hours, reflecting the thermal demands of denser AI rack configurations and continued capacity expansion.
The United States accounts for about 204 terawatt-hours of the 565 terawatt-hour global total in 2026, representing 36 percent of worldwide consumption. Within the U.S., dedicated AI data centers consume roughly 68 terawatt-hours, or one-third of the national total, while non-AI data center demand has grown only marginally over the same timeframe.
Regional electrical grids are experiencing significant strain from this growth. More than 75 data center projects valued at $130 billion were blocked in early 2026 due to opposition over power and water resource concerns. Some operators have resorted to installing on-site gas generators to activate capacity without waiting for grid connections. In Virginia, one county requested employees conserve electricity as data center demand drove utility rates higher.
AI data centers continue to become a major issue in political races around the country. Breitbart News social media director Wynton Hall has written his instant bestseller Code Red: The Left, the Right, China, and the Race to Control AI to serve as the definitive guide on how the MAGA movement can create positions on AI that benefit humanity without handing control of our nation to the leftists of Silicon Valley or allowing the Chinese to take over the world.
Read more at Tom’s Hardware here.
Lucas Nolan is a reporter for Breitbart News covering issues of AI, free speech, and online censorship.