OpenAI’s Data-Center Dream Could Consume the Power Grid
OpenAI, led by Sam Altman, has announced a strategic partnership with Nvidia to build and deploy up to 10 gigawatts of AI data centers. Nvidia could invest as much as $100 billion, a scale that could push electricity needs to rival a major city like New York at peak demand. The plan signals that AI compute is about to become a power-intensive backbone of the economy, with energy implications that ripple through grids, water supplies, and nearby communities hosting the facilities. This is more than a technology bet. It’s a test of how society will power an AI economy—and who bears the energy cost as compute grows. The announcement sits within a broader push to expand AI infrastructure, including related Stargate initiatives tied to other corporate partners.
In This Article:
Scale and Scope of the Nvidia–OpenAI Plan
The Nvidia–OpenAI plan calls for up to 10 gigawatts of AI data-center capacity, backed by as much as $100 billion in investment. Cornell energy-systems engineering professor Fengqi You notes, “Ten gigawatts is more than the peak power demand in Switzerland or Portugal. Seventeen gigawatts is like powering both countries together.” Fortune reports that Trump’s Stargate initiative could add another seven gigawatts, roughly equal to San Diego’s energy use during last year’s heat wave. OpenAI and Oracle already have a Stargate data center in Abilene, Texas, drawing enough electricity to power about half a million homes. Five new Stargate projects are expected to total seven gigawatts as part of the broader initiative.
Environmental and Infrastructure Pressures on the Grid
The scale of these plans means a major expansion of energy use and cooling needs for data centers. AI data centers require enormous amounts of electricity and water for cooling, putting pressure on local resources and power grids. If energy sources do not keep pace, carbon emissions could rise. Some observers say nuclear power may be a long-term option, though permitting and building would take years. In the near term, the industry may rely on renewables, natural gas, and retrofitting older plants to meet demand. As tens of billions are poured into infrastructure, the industry’s carbon footprint is likely to grow, especially amid a climate crisis.
Warning Voices: Clean Promise vs. Real-World Impact
'It’s scary because… now [computing] could be 10 percent or 12 percent of the world’s power by 2030,' says University of Chicago computer science professor Andrew Chien. 'They told us these data centers were going to be clean and green,' Chien told Fortune. 'But in the face of AI growth, I don’t think they can be. Now is the time to hold their feet to the fire.' Altman counterpoints with, 'Everything starts with compute' and notes that data-center infrastructure will underpin the future economy—while enabling AI breakthroughs and scalable benefits for people and businesses. The industry faces a stark choice: push for rapid innovation while taking responsibility for energy use and climate risk.
The Road Ahead: Balancing Innovation with a Sustainable Grid
The AI boom is a test of whether energy policy and technology can grow together without wrecking the climate. Analysts say energy strategy must accelerate renewables, improve efficiency, and consider long-term options like nuclear, even if those projects take years to realize. As tens of billions flow into AI infrastructure, the carbon footprint may grow unless meaningful energy transitions occur alongside compute growth. The story ends with a call to action: strengthen oversight and accountability, and ensure energy choices keep pace with innovation. Compute infrastructure is framed as the backbone of the future economy, so how we power it will shape who we are as a society.
