News: Microelectronics
21 May 2025
Infineon collaborates with NVIDIA on industry-first 800V power delivery architecture for AI data center server racks
In collaboration with NVIDIA of Santa Clara, CA, USA, Infineon Technologies AG of Munich, Germany is developing the next generation of power systems for AI data centers based on a new power delivery architecture with central power generation of 800V high-voltage direct current (HVDC).
The new system architecture significantly increases energy-efficient power distribution across the data center and allows power conversion directly at the AI chip (graphic processing unit, GPU) within the server board. Infineon says that its expertise in power conversion solutions from grid to core based on all relevant semiconductor materials silicon (Si), silicon carbide (SiC) and gallium nitride (GaN) is accelerating the roadmap to a full-scale HVDC architecture.
This step paves the way for the implementation of advanced power delivery architectures in accelerated computing data centers and will further enhance reliability and efficiency, it is expected. As AI data centers already are going beyond 100,000 individual GPUs, the need for more efficient power delivery is becoming increasingly important. AI data centers will require power outputs of 1MW and more per IT rack before the end of the decade. The HVDC architecture, coupled with high-density multi-phase solutions, will therefore set a new standard, Infineon reckons, driving the development of high-quality components and power distribution systems.
“The combination of Infineon’s application and system know-how in powering AI from grid to core, combined with NVIDIA’s world-leading expertise in accelerated computing, paves the way for a new standard for power architecture in AI data centers to enable faster, more efficient and scalable AI infrastructure,” reckons Adam White, division president Power & Sensor Systems at Infineon.
“The new 800V HVDC system architecture delivers high-reliability, energy-efficient power distribution across the data center,” says Gabriele Gorla, VP of system engineering at NVIDIA. “Through this innovative approach, NVIDIA is able to optimize the energy consumption of our advanced AI infrastructure, which supports our commitment to sustainability while also delivering the performance and scalability required for the next generation of AI workloads.”
At present, the power supply in AI data centers is decentralized, so the AI chips are supplied with power by a large number of power supply units (PSU). The future system architecture will be centralized, making the best possible use of the constraint space in a server rack. This will increase the importance of leading-edge power semiconductor solutions using fewest power conversion stages and allowing upgrades to even higher distribution voltages, says Infineon.
As a provider of power semiconductor solutions and systems integration, Infineon expects the proportion of power semiconductors in a centralized HVDC architecture to be similar or higher than in the existing AC distribution architecture. In addition to the scaling of the HVDC power architecture, Infineon continues to support state-of-the-art DCDC multi-phase solutions and intermediate architectures for hyperscalers and AI data-center operators with a broad product portfolio based on all relevant semiconductor materials along the entire power flow.
