The San Jose-based semiconductor startup is developing power technology aimed at improving energy efficiency inside AI data centers, targeting the power delivery and conversion hardware that sits between the grid and the chips. Efficient power tech at the chip level reduces energy waste and operating costs for data center operators running dense AI workloads.
Investors not disclosed
Power efficiency has become one of the hardest constraints in AI data center buildouts, with operators facing grid capacity limits and rising energy costs. A $21 million raise for a startup focused specifically on AI data center power delivery reflects growing investor conviction that semiconductor-level power solutions are needed alongside cooling and networking improvements. If the technology scales, it could reduce the energy overhead per GPU rack, a meaningful lever as hyperscalers plan facilities in the hundreds of megawatts.