Enphase Energy (ENPH) Corporate presentation summary

Market challenges for AI data centers

• AI data centers face simultaneous constraints in capacity, dynamic response, and efficiency, with demand outpacing available power infrastructure.

• U.S. data center power demand could reach ~80 GW by 2031, far exceeding projected AI-dedicated active capacity.

• AI training workloads require power infrastructure that can respond in milliseconds due to rapid load swings.

• Legacy architectures suffer from excessive power conversion stages, leading to efficiency losses.

IQ SST solution and architecture

• IQ SST is a distributed, solid-state transformer converting medium-voltage AC directly to 800 VDC, reducing sidecar infrastructure and power losses.

• Delivers 1.25 MW per unit, supports 35 kV and 15 kV interconnections, and scales up to 5 MW per skid.

• Features modular hot-swap design, five-nines availability, and no internal batteries.

• Distributed redundancy with 342 power modules per unit and 10% built-in redundancy.

• U.S.-based manufacturing and supply chain, leveraging automated, high-volume processes.

Technology and performance

• Power modules use single-stage conversion, custom Kestrel ASIC, and bi-directional GaN switching for >98.5% peak efficiency.

• Sub-millisecond response time and high-frequency switching (>250 kHz) enable rapid adaptation to dynamic loads.

• Hot-swap modularity allows service without shutdown or specialized technicians.

• Designed for reliability, with microinverter-class performance and all-plastic, potted enclosures.