• NVIDIA and Microsoft are collaborating on building a new AI supercomputer to aid enterprise customers in training and launching their own AI platforms.
• The supercomputer will make use of Microsoft’s Azure infrastructure combined with NVIDIA GPUs, networking and full stack of AI software.
• The collaboration forms part of the two companies’ wishes to increase advancements in generative AI platforms.

A new multi-year collaboration announced today will see tech giants Microsoft and NVIDIA join forces in building an AI supercomputer, expected to be one of the most powerful in the world.

The machine will reportedly be powered by Microsoft Azure’s supercomputing infrastructure combined with NVIDIA GPUs, networking and full stack of AI software, all in an effort to help enterprises train, deploy and scale their AI Products.

According to the announcement from NVIDIA, the Azure-based supercomputer will include scalable ND- and NC-series machines optimised for AI-distributed training and inference.

The machine will also include the first public cloud to incorporate NVIDIA’s advanced AI stack.

NVIDIA says this will add, “tens of thousands of NVIDIA A100 and H100 GPUs, NVIDIA QUANTUM 2 400Gb/s InfiniBand networking and the NVIDIA AI Enterprise software suite to its platform.”

It appears Microsoft and NVIDIA are looking to research and increase further advances in generative AI platforms like AI art generators DALL-E and Stable Diffusion and language model generator Megatron-Turing NLG 530B through the collaboration.

NVIDIA says this is a rapidly emerging area of AI which exist as the basis, “for unsupervised, self-learning algorithms to create new text, code, digital images, video or audio.”

To this end, the companies will also be collaborating to fine-tune Microsoft’s deep learning optimisation software DeepSpeed.

NVIDIA says that DeepSpeed will begin leveraging the NVIDIA H100 Transformer Engine, its latest Hopper architecture, to accelerate transformer-based models used for large language models, generative AI and writing computer code.

This technology is set to “dramatically accelerate” DeepSpeed’s AI calculation processes at “twice the throughput of 16-bit operations.”

“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,” said Manuvir Das, vice president of enterprise computing at NVIDIA.

“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.

[Image – Warner Bros.]