The new 42MW generator is based on the Mach 1+ jet engine. Credit: Boom Supersonic
Data center logistics are becoming harder and harder to fathom as the centralization of compute power continues to scale beyond all reasonable proportions. We've reported extensively on the incredible electrical draw of modern and future data centers, along with their astonishing level of water consumption. But data centers don't just take resources in. They also put out a major waste product: heat.
Cooling a major AI data center is tougher than it sounds, and now a solution may be coming from a company more concerned with aircraft that can break the sound barrier. Boom Supersonic is now advertising a 42MW Superpower turbine generator for powering AI data centers, and its most promising feature is its ability to continue generating power, even in hot environments.
Specifically, the Superpower generator can reportedly maintain usable output efficiency up to 43 degrees Celsius (109.4 Fahrenheit), a significant improvement over similar products that often top out at 30 degrees Celsius (86 F) or even lower. This means the generator requires less cooling and therefore uses less of its output to maintain generation.
This high heat efficiency also makes the generator "waterless," helping reduce the amount of the precious resource a data center requires.
Credit: Boom Supersonic
This new generator is based on Boom's Mach 1+ Symphony engine, designed for Boom's spiritual successor to the Concorde jet. It's aimed at resurrecting supersonic travel for regular passengers—though, as an aside, that quest has a few other problems that still need solving, too.
By remarrying their natural gas/diesel engine to a turbine generator, they can produce not thrust, but electrical power.
Boom says it has already secured an enormous purchase of 1.21GW electrical capacity from AI infrastructure company Crusoe, which should result in a sale of more than 30 of these units.
Power is the issue holding back AI from the fully integrated state the AI industry badly wants for modern life. The future of AI will require innovations in energy generation, storage, and use, with even heat efficiency becoming a major issue—whether it's jet engines, Stirling engines, or something else entirely.