Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
AI is Incredibly thirsty. at Information Center Those using these models already use huge amounts of water, and by 2030 the United States may need enough additional water capacity to match New York City’s daily water supply.
That’s according to a new study led by Shaolei Ren, associate professor of electrical and computer engineering at the University of California, Riverside, the findings of which have not yet been peer-reviewed. But it is made public on the server before printing. RX—Demonstrates that limited public water capacity has become a critical bottleneck for data center growth.
This is to avoid creating a burden on local payers. Technology companies are therefore partnering with communities to fund water infrastructure upgrades. often spending Hundreds of millions of dollars “These companies are profit driven, right? So I think something is wrong,” Ren told Gizmodo.
Why are you thirsty?
The data center is continuously operating. creating Lots of heat from the server network equipment and other forms of dense IT infrastructure, liquid cooling techniques are the most effective way to prevent overheating and system failure. But it tends to be like that. very intense.
Technology companies often argue that by using “closed-loop” cooling systems, their data centers recycle most of the water they use and minimize consumption. But even these systems can use large amounts of water. This is because many systems require an evaporative cooling tower to transfer heat outside the factory.
For example, daily water demand peaks. This is the amount needed during the hottest days of the year. For large modern data centers using evaporative cooling, this often exceeds 1 million gallons per day. And for some planned plants, it could reach 8 million gallons per day, according to the study.
water bottleneck
Public water systems are engineered to meet peak demand at all times. Therefore, the maximum water usage of a data center is an important factor in infrastructure planning. System flexibility and operational reliability. However, most operators only disclose their total annual water use. To estimate the peak water needs of U.S. data centers, Ren and his colleagues analyzed large amounts of data from public sources. Includes government records and water utility databases.
This reveals that if current water use intensity persists, U.S. data centers will need a new maximum water supply of between 697 million and 1.45 billion gallons per day by 2030, which is comparable to New York City’s typical daily water supply. Building this additional capacity could cost between $10 billion and $58 billion. with a large financial burden It falls on the community that hosts the data center..
And that’s an estimate. “Very conservative,” Ren said. His team’s calculations assume a peak-to-average daily water use ratio of just 4.5, which is on the low end of the spectrum.
This creates many problems for the technology sector. Inadequate water capacity can directly affect the feasibility and performance of data center projects. This leads to increased costs, delays and downsizing. It can also lead to operational inefficiencies. This is because data centers often need to switch to dry cooling systems. Using air instead of water When water can’t be used This is much less efficient and increases electricity demand. This puts more strain on the electrical system during the hot summer months.
Ren and his colleagues have some ideas about how to meet the growing water capacity needs of U.S. data centers. First, they emphasize the importance of requiring data centers to report peak demand. Not just total annual usage They also recommend developing partnerships between organizations and communities to find funding for infrastructure upgrades. So that residents do not have to bear the entire burden.
“I don’t see any way they can afford this type of upgrade,” Ren said. “We need funding and corporate support.”
As data centers continue to expand throughout the country The technology sector will be forced to grapple with this often overlooked bottleneck. If there are no changes These companies will face the consequences along with the communities they impact.
