Enteligent Introduces PowerCore DC-based 800VDC-to-Chip Conversion Stack for AI Data Centers

This modular three-stage rack-to-chip stack is the critical enabler for AI data center operators to confidently transition to 800 VDC distribution using less buffering hardware, less heat at the GPU and more rack space, enabling 50% greater rack compute density.

MORGAN HILL, Calif.,--(BUSINESS WIRE)--Enteligent Inc. today introduced the PowerCore™ stack, a three-stage DC/DC conversion architecture designed to deliver power from an 800 VDC AI data center rack supply to the voltage rails serving GPUs. The PowerCore stack delivers up to 50% higher compute density from the facility power to the server and includes the rack, server and GPU board, giving equipment designers a coordinated approach to high-density power delivery.

AI rack power requirements are rising sharply. Google has projected that machine-learning systems will require more than 500 kW per IT rack before 2030, while NVIDIA is developing 800 VDC infrastructure for megawatt-scale racks. As 800 VDC rapidly becomes the de facto standard for AI data center power distribution, the space required for low-voltage busbars, power shelves and transient buffering is emerging as a critical design constraint.

The PowerCore stack addresses that conversion path through three power stages:

• RackCore™ module: A 100 kW rack power shelf converts 800 VDC to a 50/48 VDC bus and eliminates the need for super capacitor and battery backup shelves which are costly and have limited life. RackCore modules enable an up to 50% higher compute density, lower costs, and higher system reliability.
• MaxCore™ module - A server-mounted module converts 50/48 VDC to a regulated 7 VDC rail. Its architecture utilizes a multi-phase, multi-path approach to higher efficiency, lower cost, and easier system implementation.
• ChipCore™ module - A compact, 16-phase module near the GPU converts the intermediate rail to 1.2/1.7 VDC, supplying the final point-of-load or vertical power stage.

The PowerCore strategy divides the overall power conversion layer into distinct electrical domains, each of which is uniquely optimized and yet exploits the advantages of adjacent stages. The result is an architecture in which voltage, current, frequency, physical distance and electromagnetic emissions at each stage of the conversion stack are considered jointly rather than independently. PowerCore takes a system-level approach to increase the overall efficiencies and reduce design complexities.

“As AI systems move toward much higher rack power, every conversion stage has to work as part of one system,” said Sean Burke, CEO of Enteligent. "The PowerCore stack draws on our DC-to-DC experience in high-density power electronics and computing to give AI infrastructure and server OEMs a modular path from 800 volts at the rack to the GPU.”

The modules can be evaluated as a complete chain or integrated individually with compatible downstream power designs. At AI-scale power levels, the optimum architecture is not necessarily the one with the fewest conversion stages; but rather that which minimizes the combined cost of conversion loss, conductor current, transient response, heat generation, electromagnetic emissions, filtering, board space and system certification from the facility bus all the way to the silicon.

Connect with the Enteligent team at the 2026 OCP Global Summit, October 12–15, in San Jose, California. To schedule a meeting, email [e-mail].

About Enteligent

Enteligent is a Silicon Valley-based developer of Direct Current (DC) power infrastructure for the world’s fastest-growing, highest-density loads, including data centers and electric vehicle charging. Built on a proprietary DC/DC platform, Enteligent’s PowerCore family addresses rack-to-chip power delivery, while its DC-coupled charging products apply direct DC-to-DC conversion to fleet charging, eliminating unnecessary AC conversions to deliver higher efficiency, lower total cost of ownership, and more scalable power architectures. To learn more, visit www.enteligent.com.