Nvidia and Microsoft will go hand in hand to shape an artificial intelligence (AI) supercomputer based on the cloud of the second, Azure. The two companies want to meet the growing demand for generative AI, like DALL-E.
Microsoft and Nvidia prepare for emerging AI models
The supercomputer will be equipped with tens of thousands of graphics processors from Nvidia, including the latest H100, released in October, among the most powerful in the world, and its predecessor, the also very powerful A100. Nvidia will also provide one of its network solutions and its AI Enterprise software suite. The latter will make the link with Microsoft’s cloud infrastructure, Azure.
This supercomputer will make it possible to accommodate generative AIs, an emerging model based on unsupervised machine learning algorithms to create new texts, codes, digital images, videos or audio content. Scott Guthrie, Executive Vice President of Cloud+ AI at Microsoft, said, Our Collaboration with NVIDIA Unlocks the World’s Most Scalable Supercomputing Platform, Bringing Cutting-Edge AI Capabilities to Every Business on Microsoft Azure “.
As part of the partnership, Nvidia will be able to take advantage of Azure’s scalable virtual machines to accelerate its progress in scalable AI. The two companies will also work on optimizing Microsoft’s DeepSpeed deep learning software.
Manuvir Das, Vice President of Enterprise Computing at Nvidia, notes that ” Technological advancements in AI as well as industry adoption are accelerating. The breakthrough in basic models has triggered a tidal wave of research, encouraged new startups and enabled new enterprise applications “.
The release date for Azure’s cloud-based AI supercomputer has not been disclosed. Anyway, Nvidia and Microsoft have left to continue this collaboration for some time: in its press release, the GPU manufacturer announces an agreement ” multi-year “.