US data centre power demand could nearly triple in the next three years, and consume as much as 12% of the country’s electricity, as the industry undergoes an artificial-intelligence transformation, reported Reuters on Friday based on a report backed by the US Department of Energy (DOE).
The Lawrence Berkeley National Laboratory produced the report as the US power industry and government attempt to understand how big tech’s data centre power demand will affect electrical grids, power bills and the climate.
Between 2017 and 2023, data centre power demand more than doubled with roll-out of more AI servers
By 2028, data centre’s annual energy consumption could reach between 74 and 132 gigawatts, or 6.7% to 12% of total US electricity consumption, according to the report.
The report included ranges that depended partly on the availability and demand for a type of AI chip known as GPUs. Currently, data centres make up a little more than 4% of the country’s power load.
“This really signals to us where the frontier is in terms of growing energy demand in the US,” said DOE’s decarbonisation office.
Surging data centre electricity needs are accompanied by rising power consumption from onshoring of US manufacturing and electrification of buildings and transportation. Overall US power demand peaked in 2024 and is expected to hit another record next year.
Reports findings may prompt DOE to upgrade the grid, including construction of long-duration battery storage at data centre sites and commercialisation of new technologies such as small nuclear reactors and advanced geothermal.
Starting in 2017, deployment of GPU-accelerated servers led to a more than doubling of the sector’s power use over a six-year period, the report said. When the last report was released in 2016, AI servers in data centres accounted for about 2% of total server energy use.
AI, which requires increasingly powerful chips and intense cooling systems, is the primary driver for the projected data centre power demand growth.
Estimates in the report are based on calculations of electricity use from installed GPUs and other data-center IT equipment, using publicly available information, data sets from market-research firms and reviews by power-sector and data centre executives.
The research team noted that knowing what’s causing the growth in energy use helps researchers to think about what opportunities there are for enhancing efficiency.
According to the report, new AI data centers are being built with power capacity as big as one gigawatt, enough to power all homes in Philadelphia.
