The Growing Demand for Data Centers in the AI Era
The next time you’re on a Zoom meeting or asking ChatGPT a question, picture this: The information zips instantaneously through a room of hot, humming servers, traveling hundreds of miles, before it makes its way back to you in just a second or two.
It can be hard to wrap your mind around, said Vijay Gadepally, a senior scientist at Massachusetts Institute of Technology’s Lincoln Laboratory, but large data centers are where nearly all artificial intelligence systems and computing happen today. ‘Each one of these AI models has to sit on a server somewhere, and they tend to be very, very big,’ he noted. ‘So if your millions or billions of users are talking to the system simultaneously, the computing systems have to really grow and grow and grow.’
As the United States aspires to become a global AI superpower, it has become home to hundreds of data centers—buildings that store and maintain the physical equipment needed to compute information.
Despite the perception that AI changes occur solely online, the rise of these technologies has tangible consequences. Data centers and the physical infrastructure required to operate them consume significant amounts of energy and resources, including water. ‘We definitely try to think about the climate side of it with a critical eye,’ expressed Jennifer Brandon, a science and sustainability consultant. ‘All of a sudden, it’s adding so much strain on the grid to some of these places.’
The Rise of Data Centers
As society has transitioned from large desktop computers to sleek laptops, the U.S. has constructed the physical infrastructure needed to support growing computing power. Large language models (LLMs) and machine learning (ML) technologies—integral to modern AI tools—have been in use for decades.
David Acosta, cofounder and chief AI officer of ARBOai, noted that while these technologies have existed for some time, they have only become widely commercialized in the past five to seven years. ‘To train and process information, these fast-learning AI models require graphic processing units (GPUs), servers, storage, cabling and other networking equipment, all housed in data centers across the country,’ he explained.
As more tasks have shifted online and computing hardware has advanced, AI models have become accessible to the general public. Training a single chatbot like ChatGPT, Acosta shared, uses about the same amount of energy as 100 homes over the course of a year. ‘And then you multiply that times the thousands of models that are being trained,’ he added. ‘It’s pretty intense.’
The U.S. currently hosts over 3,600 data centers, with approximately 80% located in just 15 states. The market has doubled since 2020, reflecting 21% year-over-year growth. Virginia is still the leader, hosting about 70% of the world’s facilities with nearly 600 centers, followed by Texas and California.
Tech companies that require substantial computing power, along with their investors, are major funders of data centers. Recently, firms including BlackRock and Microsoft invested $30 billion into new and expanded facilities, aiming for total investments of $100 billion.
Energy and Environmental Impact
Currently, data centers account for about 2% of the U.S.’ energy demand, with projections estimating this could rise to about 10% by 2027. With the development of these centers, community residents face both economic advantages and environmental challenges.
While new data centers create job opportunities and generate tax revenue, infrastructure demands raise concerns about power availability and costs. ‘New large data centers require the electricity equivalent of about 750,000 homes,’ explained a recent report. This can translate into price increases for local residents. State officials are now pushing for measures to protect consumers from rising utility bills.
Granville Martin, an environmental lawyer, highlighted local concerns, noting that communities are wary of data centers consuming available power. ‘The argument was that locals didn’t want this data center coming in there and sucking up a bunch of the available power,’ he said.
Some states are turning to alternative energy sources; Constellation Energy in Pennsylvania recently struck a deal to restart a nuclear power plant, aiming to provide carbon-free electricity to support nearby data centers. Still, cooling such industrial facilities presents additional issues. About 40% of a data center’s energy usage goes toward cooling, according to research. The infrastructure can consume as much water as that used by entire countries—for example, current AI data centers reportedly use six times the water of Denmark.
Is There a Sustainable Future for Data Centers?
The demand for energy to power AI operations is intensifying. Recent developments indicate that companies are exploring ways to improve efficiency and cut down energy consumption. Gadepally noted, ‘I think companies like DeepSeek are an example of companies doing constrained optimization.’ He emphasized that localization of AI tools could lead to significant energy savings.
As environmental concerns grow, operators might confront the question of whether they can sustain expensive, high-energy models alongside their need for efficiency. ‘You might get a better knock-knock joke from this chatbot,’ Gadepally said, ‘but that’s now using 10 times the power than it was doing before. Is that worth it?’
In conclusion, as data centers play an ever-expanding role in supporting AI advancements, their impact on energy demand and the environment raises essential questions about sustainability and responsible growth in the tech landscape.
