ECL says its hydrogen fuel cells will provide electricity to power its data centers and produce enough water to cool the servers. Credit: Weerapatkiatdumrong / Getty Images Startup ECL has emerged from stealth mode with some mighty big plans: to reinvent the data-center industry with hydrogen-powered modular data centers that use no local power and water. Rather than drawing power from the electrical grid the company will generate electricity for its data centers using hydrogen fuel cells. The only byproduct is water either as a liquid or vapor that is used for cooling with the leftovers being returned to the local environment. “So we can give back to the community some of the water that we’re producing,” said ECL founder and CEO Yuval Bachar, who previously helped design data centers for Facebook and LinkedIn. The data centers can be run like traditional colocation centers that provide the power, cooling, and racks, with customers providing their own computer hardware. The big difference will be a green option they might not be able to afford otherwise, Bachar said. ECL will also custom build their data centers to customer needs. “There’s two categories of data centers right now in the industry,” he said, “the mega data centers, and then the colocation and on-premises data centers that in most cases, are actually inferior—dramatically inferior—to the mega data centers. There’s a gap in between where ECL comes in. We are enabling the smaller customer to have the [hyhdrogen-power] experience, which they cannot have today.” The company is building its first hydrogen-powered data center in Mountain View, California, that’s due to come online in the second quarter of this year. ECLBlock architecture The ECL data-center design is modular, made up of ECLBlocks that handle the power and cooling. Each block holds up to 24 racks of servers with up to 50kW of power per cabinet, using ECL proprietary rear-door heat exchange for cooling. Bachar said the company’s sweet spot is 50kW per rack, which is more than enough for most server use cases, even power-hungry AI applications. “We can push the rack to 75 on a limited basis, not the whole data center,” he said. Customers can mix high-power racks and low-power racks without having to manually balance power across them; that is handled by AI, the company says. The company says it expects its data center is earn a 1.05 rating for power usage effectiveness (PUE), the ratio of the total energy consumed by a facility to the total energy consumed by the IT gear operating in it. Most traditional data centers have an average PUE of 1.5. Another capability of the modular design accommodates using 24 blocks to build a single large data center, for example, or partition the blocks into 24 independent facilities or any combination in between. ECL is also offering data-center-as-a-service, where it builds a data center for the customer and operates it on a consumption basis. “We actually identify the location, we get the permits, we build the structures, and we bring that to the customer,” Bachar said. The payment model is same as with a colocation facility. “So customers pay rent, and they pay for the energy,” he said. Because their facilities are off the grid, edge deployments are also candidates to take advantage of the ECL hydrogen-power technology, Bachar said. “What we’re trying to do over here is democratize it, and say, ‘Everybody can have access to—and in some cases better—technology than what the hyperscalers have,’” he said. 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