Global data center power demand is surging 26% in 2026 as artificial intelligence workloads drive unprecedented electricity consumption growth, according to Gartner Inc. Worldwide data center power consumption is projected to reach 565 terawatt hours (TWh) in 2026, up from 447 TWh in 2025, with global power demand climbing to 132 gigawatts (GW) from 104 GW the prior year.
AI-optimized servers are the primary driver of this acceleration. Gartner estimates that AI-optimized server adoption will account for 31% of data center power consumption in 2026, and by 2027, their power consumption will surpass that of conventional servers entirely. The shift reflects how compute-intensive AI workloads demand far more electricity than traditional data center operations.
Linglan Wang, Director Analyst at Gartner, noted the scale of the challenge: “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.”
In the United States, the surge is even steeper. US data center power demand is forecast to climb from 31 gigawatts in 2025 to 41 GW in 2026 and 66 GW in 2027, according to Goldman Sachs Research. This means US data centers will consume 8.5% of total peak summer power demand by 2027, nearly doubling from 4.1% in 2025, creating significant tightening across the national power market.
Power Grid Constraints Limit Deployment
The rapid growth in demand, however, is running headlong into grid limitations. Goldman Sachs Research found that only about 50-60% of data center capacity scheduled for the next one to two years is expected to come online on time. Supply chain delays, labor shortages, and the typical 18-to-24-month construction timeline once permits are secured all contribute to the gap between planned and realized capacity.
Regional impacts will diverge sharply. The Mid-Atlantic, Mid-Continent, and Northwest markets face elevated reliability risks because their planned generation capacity additions are limited relative to incoming data center demand. These regions may ultimately have to turn some data centers away. Texas and Georgia, by contrast, have significant new power generation capacity in the pipeline and should see only marginal tightening. Meanwhile, Tennessee, New England, and Florida are already so constrained that data center additions will likely be severely limited.
Power availability and time-to-client have become the primary factors driving where data centers choose to locate. Tech companies building AI infrastructure are no longer choosing based on labor costs or real estate prices alone—they are choosing based on where power can reach them fastest. This shift reflects how AI infrastructure investment decisions now hinge entirely on electricity access.
Gartner projects that data center power consumption will reach over 1,200 terawatt hours by 2030, far exceeding what the grid can currently supply. The firm warned that infrastructure and operations leaders must prioritize efficiency upgrades, secure grid access, and invest in high-efficiency cooling systems and edge computing to mitigate power constraints and ensure sustainable growth.
Sources
- Gartner — Global data center electricity consumption forecast to grow 26% in 2026 to 565 TWh; AI-optimized servers to account for 31% of power consumption; projection of 1,200 TWh by 2030.
- Goldman Sachs Research — US data center power demand forecast to climb from 31 GW in 2025 to 41 GW in 2026 and 66 GW in 2027; only 50-60% of scheduled capacity expected to come online on time; regional impacts and reliability risks.
