Nautilus Data Technologies brings the experience of cooling nuclear power plants to cooling data center floating on barges in bodies of water. Credit: romolotavani / vladimir timofeev / Getty Images Data centers and water seem to go together, despite the fact water is bad for electronics. Many hyperscale data centers are built near rivers for use as hydroelectric power sources, liquid cooling is growing in popularity, and in one extreme case, Microsoft sunk a mini data center off the coast of northern England. The next step, it seems, is the floating data center, one on the water and easily accessible but also mobile. A startup called Nautilus Data Technologies has lined up $100 million in funding to build a six-megawatt floating colocation facility that it says will be cheaper and more efficient than traditional facilities. Based in Pleasanton, California, Nautilus has invented what it calls the TRUE (Total Resource Usage Effectiveness) cooling system that integrates maritime and industrial cooling technologies with data center infrastructure to increase the energy efficiency of its operations, reduce the overall cost of computing, and decrease greenhouse gases and air pollution. Now while the idea of water cooling is hardly new, I was a little flummoxed at Nautilus’s strategy, especially since its first data center will be based in Stockton, California, a city repeatedly voted one of the worst places to live, and the Calaveras River that runs through the town is filthy. There’s a method to the madness, though. James Connaughton, president and CEO of Nautilus, comes to data center cooling from the industrial cooling industry. He’s used to cooling factories, petrochemical plants, and nuclear power plants. Servers for him are child’s play. “If you can cool a nuclear power plant generating a gigawatt of heat, 20 megawatts is nothing,” he told me. Nautilus employs the well understood method of once-through cooling that is used in every other heat-generating sector except data centers. It takes in water, filters it for solids, cycles it past a heat exchanger and returns it to the river or sea a little warmer than it came in but otherwise unchanged. Inside the floating data center is a closed loop of pure water used for the rear door exchanges and direct-to-chip cooling. Nautilus’s secret sauce is a transfer of coolness from the outside water source to the internal closed loop. The outside water cools the internal loop, absorbs the heat, and is dumped back in the river. This eliminates the old methods that require evaporative chilling, which means using refrigerants and generating waste. “We’re not harming the environment on the water side, and on the closed loop side we can control it to high standards,” says Connaughton. Because it’s about heat rejection, the water needs to be cool not cold since it’s all relative to server temperature. All that matters is that the river water is colder than the closed loop water. So they can do it in Stockton, which gets pretty hot in the summer, regardless of seasonal weather. Connaughton said they could put a barge in Lake Mead just outside of Las Vegas if they wanted to. “We can have a constant [power-usage efficiency] regardless of weather,” he says. The Stockton barge would be modest in size compared to most data centers, but Connaughton says there is no limit in size at all. The Stockton barge will establish the template and can scale up in size. “With our first project, we wanted to prove three things: the feasibility of once-through cooling, doing it in a chemical-free way, and show an option for full pre-manufactured data center, which the barge does,” he says. The Stockton data center will be commissioned in the fourth quarter of this year. Related content news Supermicro unveils AI-optimized storage powered by Nvidia New storage system features multiple Nvidia GPUs for high-speed throughput. By Andy Patrizio Oct 24, 2024 3 mins Enterprise Storage Data Center news Nvidia to power India’s AI factories with tens of thousands of AI chips India’s cloud providers and server manufacturers plan to boost Nvidia GPU deployment nearly tenfold by the year’s end compared to 18 months ago. By Prasanth Aby Thomas Oct 24, 2024 5 mins GPUs Artificial Intelligence Data Center news Gartner: 13 AI insights for enterprise IT Costs, security, management and employee impact are among the core AI challenges that enterprises face. By Michael Cooney Oct 23, 2024 6 mins Generative AI Careers Data Center news Network jobs watch: Hiring, skills and certification trends What IT leaders need to know about expanding responsibilities, new titles and hot skills for network professionals and I&O teams. By Denise Dubie Oct 23, 2024 33 mins Careers Data Center Networking PODCASTS VIDEOS RESOURCES EVENTS NEWSLETTERS Newsletter Promo Module Test Description for newsletter promo module. Please enter a valid email address Subscribe