TI and NVIDIA

TI and NVIDIA to bring adept power dole out to AI infra

Share

Dallas, USA: Texas Instruments (TI) and NVIDIA are working together in the development of power management and sensing technologies.

TI and NVIDIA partnering

TI said the power management and sensing technologies development will help enable NVIDIA’s future 800V high-voltage direct current (HVDC) power distribution systems for data centre servers. The new power architecture paves the way for more scalable and reliable next-generation AI data centres.

With the growth of AI, the power required per data centre rack is predicted to increase from 100kW today to more than 1MW in the near future. To power a 1MW rack, today’s 48V distribution system would require almost 450lbs of copper, making it physically impossible for a 48V system to scale power delivery to support computing needs in the long term.

The new 800V high-voltage DC power-distribution architecture will provide the power density and conversion efficiency that future AI processors require, while minimising the growth of the power supply’s size, weight and complexity. This 800V architecture will enable engineers to scale power-efficient racks as data centre demand evolves.

“A paradigm shift is happening right in front of our eyes,” said Jeffrey Morroni, Kilby Labs’ Director of Power Management Research and Development and a TI Fellow.

“AI data centres are pushing the limits of power to previously unimaginable levels. A few years ago, we faced 48V infrastructures as the next big challenge. Today, TI’s expertise in power conversion combined with NVIDIA’s AI expertise are enabling 800V high-voltage DC architectures to support the unprecedented demand for AI computing,” added Morroni.

“Semiconductor power systems are an important factor in enabling high-performance AI infrastructure,” said Gabriele Gorla, VP of System Engineering of NVIDIA.

“NVIDIA is teaming with suppliers to develop an 800V high-voltage DC architecture that will efficiently support the next generation of powerful, large-scale AI data centres,” added Gorla.