Microsoft partners up with tech firm to build supercomputer

Microsoft and Nvidia have announced a collaboration to build a world-class supercomputer driven by artificial intelligence (AI) in a project expected to span several years.

The new machine will combine tens of thousands of graphics processing units (GPUs) designed by Nvidia with Microsoft’s cloud-based supercomputing network Azure and is intended to help further the uptake of AI across industries at the enterprise level.

“AI technology advances as well as industry adoption are accelerating,” said Manuvir Das, vice president of enterprise computing at Nvidia. “The breakthrough of foundation models has triggered a tidal wave of research, fostered new startups, and enabled new enterprise application. Our collaboration with Microsoft will provide researchers and companies with state-of-the-art AI infrastructure and software to capitalize on the transformative power of AI.”

“AI is fueling the next wave of automation across enterprises and industrial computing, enabling organizations to do more with less as they navigate economic uncertainties,” said Scott Guthrie, executive vice president of the Cloud + AI Group at Microsoft. “Our collaboration with Nvidia unlocks the world’s most scalable supercomputer platform, which delivers state-of-the-art AI capabilities for every enterprise on Microsoft Azure.”

The project will also augment Microsoft’s DeepSpeed machine learning software by adapting Nvidia’s AI software development kits to work with the Azure system. The Azure network will be the first cloud-based instance to adopt Nvidia’s quantum-based InfiniBand networking at a speed of 400 gigabits per second – doubling the rate it currently runs on.

This will enable corporate clients to use thousands of GPUs in single clusters to train large-scale AI models that can take on workloads of any size.