Nvidia and Microsoft announced a joint effort to build one of the most powerful artificial intelligence (AI) supercomputers in the world. The goal is to create a machine that can handle the massive computational workloads required to train and scale artificial intelligence (AI) applications. Nvidia will use Microsoft’s scalable virtual machine instances to accelerate developments in AI models like DALL-E.
See also: Microsoft confirms gaming performance issues in Windows 11 22H2
Nvidia’s powerful H100 and A100 data center GPUs, combined with the Quantum-2 InfiniBand networking platform, will form the foundation of the AI supercomputer built on Microsoft’s Azure cloud infrastructure. According to Nvidia, the combination of Microsoft’s Azure cloud platform and Nvidia’s GPUs , networking, and full AI suite will enable more enterprises to train, deploy, and scale AI — including large, state-of-the-art models. Finally, part of the agreement between the two companies includes a commitment to jointly develop DeepSpeed — Microsoft’s deep learning optimization software.
See also: Microsoft fixes Windows DirectAccess connectivity issues
In a statement, Nvidia said the supercomputer could be used to “research and further accelerate developments in Generative AI,” which refers to self-learning algorithms that generate a wide range of content, such as text, code, digital images, video or audio. These artificial intelligence models have seen rapid growth in recent years — accompanied by increased demand for powerful computing infrastructure capable of scaling along with development .
“The development and application of AI technology is growing rapidly,” said Manuvir Das, vice president of enterprise computing at Nvidia. “New discoveries in fundamental models have given way to a wealth of new research opportunities, startups , and enterprise applications . With Microsoft, we will be able to provide researchers and enterprises with access to the best AI infrastructure and software available, so they can fully harness this transformative power.”
Source of information: theverge.com
