AI Data Center Boom Accelerates Shift to Solid-State Transformers

AI-generated image · US National Wire
New semiconductor-based power hardware aims to bypass the manufacturing bottlenecks of industrial-age grid infrastructure.
The surge in electricity demand from AI data centers is straining U.S. power grids, but according to reporting from Ars Technica, this boom is accelerating the deployment of solid-state transformers.
Traditional transformers rely on a design from the 1880s, featuring copper coils wound around steel cores. Because these units are custom-built and manually assembled, utility companies often face delivery delays spanning several years, which Ars Technica reports hinders both grid expansion and the replacement of aging equipment.
Solid-state transformers replace this manual process with high-frequency semiconductor switching using materials such as silicon carbide. This allows the units to be mass-manufactured, resulting in hardware that is lighter, smaller, and modular. Srdjan Lukic, a professor of electrical and computer engineering at North Carolina State University, told Ars Technica that data centers are the "killer application" for this tech because they utilize direct current (DC) power architectures. Solid-state transformers can convert alternating current (AC) from the grid directly into DC power, eliminating the need for separate conversion devices and reducing the necessary infrastructure and wiring.
Commercial interest is scaling rapidly. Ars Technica notes that DG Matrix, Heron Power, and Amperesand have collectively raised over $280 million in the last year to bring this technology to market. Beyond data centers, the technology has applications for EV charging. A collaboration involving the New York Power Authority and NC State recently tested a 1 MW solid-state transformer system at an Electric Power Research Institute site in Lenox, Massachusetts, demonstrating its ability to handle voltage step-down and AC-to-DC conversion for vehicle batteries.

