MENU

Infineon and NVIDIA partner on power delivery for AI data centres

Infineon and NVIDIA partner on power delivery for AI data centres

Technology News |
By Jean-Pierre Joosting



Infineon Technologies AG, in collaboration with NVIDIA, is developing the next generation of power systems based on a new architecture featuring central power generation at 800 V and high-voltage direct current (HVDC) for power distribution within AI data centres.

Significantly enhancing energy-efficient power distribution across AI data centres, the new system architecture enables power conversion directly at the AI chip (GPU) within the server board. Infineon is leveraging its expertise in power conversion from the grid to the core, based on all relevant semiconductor materials, including silicon (Si), silicon carbide (SiC), and gallium nitride (GaN), to drive the roadmap towards a full-scale HVDC architecture.

As AI data centres are already utilising more than 100,000 individual GPUs, the need for more efficient power delivery is becoming increasingly important. It is estimated that AI data centres will require power outputs of one megawatt or more per IT rack by the end of the decade. Therefore, the HVDC architecture, coupled with high-density multiphase systems, will set a new standard for the industry, driving the development of high-quality components and power distribution systems.

“Infineon is driving innovation in artificial intelligence,” said Adam White, Division President of Power and Sensor Systems at Infineon. “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 centres to enable faster, more efficient and scalable AI infrastructure.”

“The new 800V HVDC system architecture delivers high reliability, energy-efficient power distribution across the data centre,” said Gabriele Gorla, vice president of system engineering at NVIDIA. “Through this innovative approach, NVIDIA can optimise 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.”

Currently, the power supply in AI data centres is decentralised. This means that the AI chips are supplied with power by a large number of power supply units. The future system architecture will be centralised, making the best possible use of the constraint space in a server rack. This will increase the importance of leading-edge power semiconductor solutions that use the fewest power conversion stages and allow upgrades to even higher distribution voltages.

 

Key benefts of the HVDC architecture

Currently, power distribution in AI data centres involves multiple voltage conversions, which, as power requirements increase due to the computational demands of AI, lead to growing inefficiencies and complexities. The NVIDIA 800 V HVDC architecture converts high-voltage AC grid power directly to 800 V HVDC at the data centre perimeter using industrial-grade rectifiers. This eliminates most intermediate conversion steps, along with associated energy losses during numerous AC/DC and DC/DC transformations. 

The HVDC architecture also significantly reduces the number of power supply units with fans required in the power chain, enabling higher system reliability, lower heat dissipation, a reduced component count, and improved energy efficiency. As cooling is a significant portion of the energy budget in AI data centres, reducing heat dissipation has a substantial impact on equipment and energy costs. Maintenance costs can also be reduced by up to 70% due to fewer power supply unit failures.

Using 800 V busways instead of AC in electrical distribution enables 85% more power to be transmitted through the same conductor size, as the higher voltage reduces current demand and lowers resistive losses. With lower current, thinner conductors can handle the same load. This reduces copper requirements by 45% and eliminates AC-specific inefficiencies, including skin effects and reactive power losses.

Reference: NVIDIA 800 V HVDC Architecture Will Power the Next Generation of AI Factories.

www.infineon.com
www.nvidia.com

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News

Share:

Linked Articles
10s