The announcement includes new offerings on Microsoft Azure cloud and 50+ new partner systems for accelerating scientific discovery.
NVIDIA announced broad adoption of its next-generation H100 Tensor Core GPUs and the Quantum-2 InfiniBand architecture, including new offerings on Microsoft Azure cloud and over 50 new partner systems "for accelerating scientific discovery."
At SC22, the company also released major updates to its cuQuantum, CUDA, and BlueField DOCA acceleration libraries and announced support for Omniverse on NVIDIA A100- and H100-powered systems.
The updates are part of NVIDIA’s HPC platform – a full technology stack with CPUs, GPUs, DPUs, systems, networking, and a broad range of AI and HPC software. It enables researchers to accelerate their work on powerful systems, on-premises or in the cloud.
“AI is reinventing the scientific method. Learning from data, AI can predict impossibly complex workings of nature, from the behavior of plasma particles in a nuclear fusion reactor to human impact on regional climate decades in the future,” said Jensen Huang, founder and CEO of NVIDIA. “By providing a universal scientific computing platform that accelerates both principled numerical and AI methods, we’re giving scientists an instrument to make discoveries that will benefit humankind.”
Among NVIDIA’s partners announcing H100-powered servers are ASUS, Atos, Dell Technologies, INGRASYS, GIGABYTE, Hewlett Packard Enterprise, Lenovo, Penguin Solutions, QCT, and Supermicro. Every H100 PCIe GPU includes a five-year license for NVIDIA AI Enterprise, a cloud-native software suite that streamlines the development and deployment of AI.
Take a look at the updates to acceleration libraries:
- NVIDIA CUDA libraries now include a multi-node, multi-GPU Eigensolver enabling unprecedented scale and performance for leading HPC applications like VASP, a package for first-principles quantum mechanical calculations.
- The NVIDIA cuQuantum software development kit for accelerating quantum computing workflows now supports approximate tensor network methods. This allows researchers to simulate tens of thousands of qubits, as well as automatically enables multi-node, multi-GPU support for quantum simulation with unparalleled performance using the cuQuantum Appliance.
- NVIDIA DOCA, the open cloud SDK and acceleration framework for NVIDIA BlueField DPUs, includes advanced programmability, security and functionality to support new storage use cases.
In addition, NVIDIA announced that Omniverse can be used by the scientific computing community on NVIDIA A100 and NVIDIA H100 systems for a variety of workloads.
Find the details here and don't forget to join our Reddit page and our Telegram channel, follow us on Instagram and Twitter, where we share breakdowns, the latest news, awesome artworks, and more.