Scalable IQ GaN-Based Solid-State Transformer Architecture Eyes AI Data Centers

Enphase’s new SST architecture, which aligns with NVIDIA’s 800-V DC roadmap, uses a meshed array of sub-1-kV bidirectional GaN switches to provide scalability and distributed redundancy.

What you’ll learn:

  • Enphase is developing the IQ Solid-State Transformer (IQ SST), a distributed solid-state transformer platform purpose-built for AI data centers.
  • The IQ SST architecture differs from most other SST designs because it’s based on a scalable, redundant array of modules based on medium-voltage bidirectional GaN switches.

Enphase Energy recently announced that it has begun development of the IQ Solid-State Transformer (IQ SST), a distributed solid-state transformer platform purpose-built for AI data centers. It’s intended to replace traditional centralized power conversion with a distributed “supercluster” architecture.

IQ SST will convert medium-voltage AC, including 35- and 15-kV interconnection classes, directly into regulated DC power in a single conversion stage. The platform is designed to support both 800- and ±400-V DC rack configurations defined by emerging AI data center standards, with sub-millisecond response to dynamic AI loads. The new design is aligned with NVIDIA’s 800-V DC roadmap (Rubin Ultra 2027, Feynman 2028) and OCP’s Mt. Diablo ±400-V architecture.

According to Enphase, each 1.25-MW IQ SST rack will combine 342 intelligent, semiconductor- and software-defined power modules operating in a coordinated series/parallel configuration. Their built-in redundancy enables continued operation with only 90% of the power modules, which is expected to deliver 98.5% efficiency and 99.999% availability.

The IQ SST architecture differs from traditional transformer-based power solutions, and other SST designs, in other important ways:

  • Single-stage conversion straight from medium-voltage AC to 800 V DC. Collapses several conversion stages and gets the sidecar out of the white space.
  • It offers sub-millisecond control response, thanks to it being built on the proprietary Kestrel ASIC (22 nm), designed to enable fast, coordinated response across hundreds of distributed power modules.
  • It uses bidirectional GaN switches under 1 kV, sourced from a widely available supply base instead of the high-voltage silicon-carbide (SiC) stacks being used by some monolithic SST designs.

The IQ SST platform, designed without internal batteries, is intended to reduce or eliminate the need for rack-level battery sidecars and traditional uninterruptible-power-supply (UPS) systems in supported data center configurations. For data center operators, that means less power infrastructure, less floor space consumed by electrical equipment, and more space available for GPUs, thereby enabling capacity expansion within existing facilities.

AI data centers are the first target application for the platform, as they rapidly migrate toward 800- and ±400-V DC power architectures. Over time, Enphase expects to extend the same distributed, semiconductor- and software-defined powe- conversion architecture into larger adjacent high-power markets.

Full system demonstrations of the IQ SST platform are scheduled for late this year, with customer pilots in 2027 and volume shipments anticipated in 2028. To learn more about the Enphase solid-state transformer solution, visit the Enphase website.

Enphase has also published a technical white paper titled “Intelligent Power for AI,” which provides technical details of the IQ SST architecture.

About the Author

Lee Goldberg

Contributing Editor

Lee Goldberg is a self-identified “Recovering Engineer,” Maker/Hacker, Green-Tech Maven, Aviator, Gadfly, and Geek Dad. He spent the first 18 years of his career helping design microprocessors, embedded systems, renewable energy applications, and the occasional interplanetary spacecraft. After trading his ‘scope and soldering iron for a keyboard and a second career as a tech journalist, he’s spent the next two decades at several print and online engineering publications.

Lee’s current focus is power electronics, especially the technologies involved with energy efficiency, energy management, and renewable energy. This dovetails with his coverage of sustainable technologies and various environmental and social issues within the engineering community that he began in 1996. Lee also covers 3D printers, open-source hardware, and other Maker/Hacker technologies.

Lee holds a BSEE in Electrical Engineering from Thomas Edison College, and participated in a colloquium on technology, society, and the environment at Goddard College’s Institute for Social Ecology. His book, “Green Electronics/Green Bottom Line - A Commonsense Guide To Environmentally Responsible Engineering and Management,” was published by Newnes Press.

Lee, his wife Catherine, and his daughter Anwyn currently reside in the outskirts of Princeton N.J., where they masquerade as a typical suburban family.

Lee also writes the regular PowerBites series

Sign up for our eNewsletters
Get the latest news and updates

Comment About the Article

To join the conversation, and become an exclusive member of Electronic Design, create an account today!