Innovation in IT got us thus far. Graphics processing units (GPUs) that power the computing behind AI have fallen in cost by 99% since 2006. There was similar concern concerning the energy use of information centers within the early 2010s, with wild projections of growth in electricity demand. But gains in computing power and energy efficiency not only proved these projections unsuitable but enabled a 550% increase in global computing capability from 2010 to 2018 with only minimal increases in energy use.
Within the late 2010s, nonetheless, the trends that had saved us began to interrupt. Because the accuracy of AI models dramatically improved, the electricity needed for data centers also began increasing faster; they now account for 4.4% of total demand, up from 1.9% in 2018. Data centers devour greater than 10% of the electricity supply in six US states. In Virginia, which has emerged as a hub of information center activity, that figure is 25%.
Projections concerning the future demand for energy to power AI are uncertain and range widely, but in a single study, Lawrence Berkeley National Laboratory estimated that data centers could represent 6% to 12% of total US electricity use by 2028. Communities and firms will notice one of these rapid growth in electricity demand. It can put pressure on energy prices and on ecosystems.The projections have resulted in calls to construct numerous recent fossil-fired power plants or bring older ones out of retirement. In lots of parts of the US, the demand will likely lead to a surge of natural-gas-powered plants.
It’s a frightening situation. Yet once we zoom out, the projected electricity use from AI continues to be pretty small. The US generated about 4,300 billion kilowatt-hours last yr. We’ll likely need one other 1,000 billion to 1,200 billion or more in the following decade—a 24% to 29% increase. Almost half the extra electricity demand might be from electrified vehicles. One other 30% is predicted to be from electrified technologies in buildings and industry. Innovation in vehicle and constructing electrification also advanced within the last decade, and this shift might be excellent news for the climate, for communities, and for energy costs.
The remaining 22% of latest electricity demand is estimated to return from AI and data centers. While it represents a smaller piece of the pie, it’s essentially the most urgent one. Due to their rapid growth and geographic concentration, data centers are the electrification challenge we face right away—the small stuff we have now to determine before we’re in a position to do the massive stuff like vehicles and buildings.
We also need to know what the energy consumption and carbon emissions related to AI are buying us. While the impacts from producing semiconductors and powering AI data centers are necessary, they’re likely small compared with the positive or negative effects AI could have on applications corresponding to the electricity grid, the transportation system, buildings and factories, or consumer behavior. Corporations could use AI to develop recent materials or batteries that might higher integrate renewable energy into the grid. But they may also use AI to make it easier to seek out more fossil fuels. The claims about potential advantages for the climate are exciting, but they have to be repeatedly verified and can need support to be realized.
This isn’t the primary time we’ve faced challenges coping with growth in electricity demand. Within the Nineteen Sixties, US electricity demand was growing at greater than 7% per yr. Within the Seventies that growth was nearly 5%, and within the Nineteen Eighties and Nineties it was greater than 2% per yr. Then, starting in 2005, we principally had a decade and a half of flat electricity growth. Most projections for the following decade put our expected growth in electricity demand at around 2% again—but this time we’ll should do things in a different way.
To administer these recent energy demands, we’d like a “Grid Latest Deal” that leverages private and non-private capital to rebuild the electricity system for AI with enough capability and intelligence for decarbonization. Latest clean energy supplies, investment in transmission and distribution, and techniques for virtual demand management can cut emissions, lower prices, and increase resilience. Data centers bringing clean electricity and distribution system upgrades may very well be given a quick lane to hook up with the grid. Infrastructure banks could fund recent transmission lines or pay to upgrade existing ones. Direct investment or tax incentives could encourage clean computing standards, workforce development within the clean energy sector, and open data transparency from data center operators about their energy use in order that communities can understand and measure the impacts.
