Supersonic airline outfit Boom unveils turbine for AI data centers — 42 MW Superpower turbine uses the same tech designed to power Concorde successor to Mach 1.7 at 60,000 ft
Besides having already eaten most sorts of chip production, AI is quickly coming for power sources. Gas turbines are seeing increased demand, and companies like xAI are using tens of units to power their data centers. Now, Boom, known for its supersonic airliner project, has revealed the Superpower turbine with a bang, a new project that will repurpose its aircraft engine tech to power AI data centers.
The Superpower turbine is purportedly "optimized for AI datacenters," a claim boiled down to delivering its full 42 MW with 39% efficiency, at an operating temperature up to a toasty 110° F (43° C). Judging by a quick search, that's significantly higher than contemporary designs, whose output drops at around 86° F (30° C), if not much sooner—and precisely when it's most needed, when servers are working the hardest.
Almost as importantly, Boom claims the new design is "waterless", a fact that should somewhat assuage ever-growing concerns about datacenters' water usage. Although the servers' cooling towers still need water to function, the fact that it would no longer be necessary to cool the turbines would definitely be a bonus.
Boom says it's taken an order for its new design from Crusoe AI for a substantial 1.21 GW of turbines. The company says it builds everything in Denver, Colorado, and that it intends to create a Superfactory for this new enterprise. Boom expects to deliver 200 MW worth of turbines by 2027, 1 GW in 2028, and up to a meaty 2 GW come 2029. Crusoe ordered 29 units, and the first of them ought to be delivered in 2027.
According to Boom's Blake Scholl, the idea came about due to the fact that xAI and OpenAI both are using converted jet engines, whose main limitation is twofold: commercial availability and falling output at high ambient temperatures, as they're designed for high-altitude operation. The Superpower turbine design is derived from the company's existing Symphony airspace-targeted turbine.
According to Boom, the Symphony engine is designed to power supersonic flight at speeds of up to Mach 1.7 at 60,000 feet. Superpower will run on natural gas, rather than jet fuel, of course. While news of a Concorde successor pivoting its tech to AI might give aviation fans pause, Boom says the new project will actually boost its core mission. "The fastest way to a certified, passenger-carrying Symphony engine is to run its core for hundreds of thousands of hours in the real world, powering Earth’s most demanding AI data centers," Boom says.
Turns out these supersonic jet-powered AI data centers will actually serve as a giant test bed for Boom's turbines. Naturally a profitable venture, Boom also says that the profitability from Superpower will fund "the remainder of the aircraft program," creating a self-sustaining path to supersonic flight.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
It can be argued that the long-term future for energy generation is nuclear and with more renewables. The AI train stops for nobody, though, so having better local power generation options now, with rising power costs and stressed grids, is as good a stopgap as any.
Bruno Ferreira is a contributing writer for Tom's Hardware. He has decades of experience with PC hardware and assorted sundries, alongside a career as a developer. He's obsessed with detail and has a tendency to ramble on the topics he loves. When not doing that, he's usually playing games, or at live music shows and festivals.
I guess that's one way of acquiring funds (est. $7~20bn) for a highly ambitious project.
Certifying an aircraft takes forever, and those Symphony engines don't meet the original goals. They lack range, speed, and fuel efficiency, so I find it funny they are pitching the engines as "fuel efficient".
I wish them the best of luck, because the technical side of the jet is quite interesting, even if there's nothing revolutionary about the design. It's all proven technology, but in a modern package.
415zXv8u3rI
Think of all the things that could be done with the power from that order of turbines. 1.21 Gigawatts.... That's just enough to run your Flux Capacitor
39% efficiency is pretty bad. How about pump the brakes on building data centers until they build the small nuclear plants to power them.
It's normal percentage when have not cogeneration but simple gases rotating turbine and generator. So many energy in the form of the heat just is ejected in atmosphere.
