Technology companies have invested so much money in construction data centers active in recent months driving force of the US economy— and the AI race shows no signs of slowing down. Meta chief Mark Zuckerberg told President Donald Trump last week that the company would spend $600 billion on US infrastructure, including data centers, by 2028, while OpenAI has already committed to spending $1.4 trillion.
A sweeping new analysis looks at the environmental impacts of data centers in the US to understand exactly what the country could face as this construction continues over the next few years, and where the US should build data centers to avoid the most harmful environmental impacts.
studypublished in the journal Nature Communications on Monday, uses a variety of data, including demand for artificial intelligence chips and information on state electricity and water shortages, to predict the potential environmental impact of future data centers before the end of the decade. The study models a number of different possible scenarios for how data centers could impact the US and the planet, and warns that tech companies' net-zero promises are unlikely to keep up with the energy and water demands of the massive facilities they build.
Fengqi Yu, a professor of energy systems engineering at Cornell and one of the authors of the analysis, says the research, which began three years ago, is “an ideal time to understand how AI impacts climate systems and water use and consumption.”
The artificial intelligence industry is “growing much faster than we expected,” he adds, especially given the Trump administration's intense focus on the industry. “It’s all gaining so much momentum now.”
Not all data centers are created environmentally equal: a significant portion of their water and carbon footprint depends on where they are located. Some US states may have grids that run more on renewable energy or are making great strides in bringing cleaner energy to the grid; this significantly reduces the carbon emissions of data centers that draw power from these grids. Likewise, states with less water stress are better positioned to provide the large quantities of water needed to cool data centers. (Cooling also accounts for a significant portion of data center energy consumption.) The best places to locate data centers in the U.S. over the next few years will be states that strike a balance between these two resources: Texas, Montana, Nebraska and South Dakota are “optimal candidates for installing artificial intelligence servers,” the analysis shows.
Most data center construction in the US has historically been concentrated in places such as Virginia, the US data center, and Northern California. Proximity to Washington, D.C. and Silicon Valley was important to data center companies, as was the dense fiber optic connectivity in those regions and their skilled workforce. Virginia has also for years offered significant tax incentives for data centers, one of the methods other states have resorted to to encourage development. According to Data center mapan industry tool that tracks data center development, of the more than 4,000 data centers in the U.S., more than 650 are in Virginia (the most in the country), and California has more than 320, ranking third.
