"The future of transformative enterprise technologies such as AI and HPC is in next-generation cloud platforms like Microsoft Azure, where innovators have the opportunity to deliver a new era of technological breakthroughs," said Microsoft general manager of Azure AI Infrastructure Nidhi Chappell.

"The Nvidia Quantum-2 InfiniBand networking platform equips Azure with the throughput capabilities of a world-class supercomputing centre, available at cloud scale and on demand, and allows researchers and scientists using Azure to achieve their life's work."

One of the new H100 based systems is the Dell PowerEdge XE9680, aimed at demanding AI and high-performance workloads.

"AI is propelling innovation unlike any technology before it," said Dell Technologies vice president of portfolio and product management for PowerEdge, HPC and core compute at Dell Technologies Rajesh Pohani.

"Dell PowerEdge servers with Nvidia Hopper GPUs support customers to push the boundaries and make possible new discoveries across industries and institutions."

Nvidia has also released major updates to its cuQuantum, Cuda and BlueField Docatm acceleration libraries, and announced support for its Omniverse simulation platform on NVIDIA A100 and H100 based systems.

These updates variously include a multi-node, multi-GPU Eigensolver; support for approximate tensor network methods; and support for new storage use cases.

In addition, the Nvidia Omniverse platform for metaverse applications now connects to scientific computing visualisation software and supports new batch-rendering workloads on systems powered by A100 and H100 Tensor Core GPUs, while Nvidia OVX and Omniverse Cloud together allow the development of fully real-time scientific and industrial digital twins for the high performance computing community.

Omniverse now supports batch AI and HPC workloads – including rendering videos and images or generating synthetic 3D data – on existing A100 or H100 systems.

Nvidia also provides connections to scientific computing tools such as Kitware's ParaView, Nvidia IndeX, Nvidia Modulus, and NeuralVDB.

"Today's scientific computing workflows are extremely complex, involving enormous datasets that are impractical to move and large, global teams that use their own specialised tools," said Nvidia lead product manager of accelerated computing Dion Harris.

"With new support for Omniverse on A100 and H100 systems, HPC customers can finally start to unlock legacy data silos, achieve interoperability in their complex simulation and visualisation pipelines, and generate compelling visuals for their batch-rendering workflows."