The most powerful supercomputers on earth are used to perform all kinds of complex operations. Increasingly, they are being used to enable artificial intelligence for research that could one day affect billions of people.
The world’s fastest and most powerful high-performance computing (HPC) supercomputers are at the forefront of the International Supercomputing Conference (ISC), which runs from May 29 to June 2 in Hamburg, Germany. As part of the ISC event, Nvidia will provide insights into its latest HPC systems and cases of their efficient use.
“HPC Plus AI is indeed a transformative tool for scientific computing,” said Dion Harris, Chief Technical Product Marketing Manager for Accelerated Computing, at a media briefing ahead of the ISC. “We’re talking about exascale AI because we believe it will be a key tool for driving scientific innovation, and any data center building a supercomputer needs to understand how their system will work from an AI perspective.”
US-based Grace Hopper Superchip Supercomputer arrives at Los Alamos National Laboratory
Nvidia first announced its Grace ARM-based CPU in April 2021, with the goal of incorporating them into HPC deployments. The goal is now being realized.
At ISC 2022, Nvidia is announcing that Los Alamos National Laboratory and Hewlett-Packard Enterprise (HPE) are building Venado, the first US-based supercomputer to use grace chip architecture.
The Vanado supercomputer uses a combination of Grace and Grace Hopper superchips in a system that is expected to provide 10 exoflops of AI performance. The Venado system will be used for research in physics, renewable energy as well as energy distribution.
Nvidia-powered AI enables brain imaging research
Brain imaging HPC and AI use cases are being announced by Nvidia in ISC 2022.
King’s College London is using the Nvidia-powered Cambridge-1 system, the most powerful supercomputer in the United Kingdom, along with the open-source Monai AI framework that has been optimized for medical imaging use cases.
Powerful hardware and AI software have been used to create the world’s largest database of artificial brain images.
Harris explained that the amount of AI-driven research is important in identifying conditions such as Alzheimer’s or dementia. “But to train those models, you need a big database,” he said.
There are many privacy concerns when using actual patient data, which is why it is important for researchers to have access to synthetic data, he added.
“This is a true example of an HPC that not only speeds up and distributes feeds, but also makes a real contribution to the scientific and research community,” Harris said.
Modeling of nuclear fusion reactors
As people around the world try to find solutions to the challenges of global warming, one of the primary strategies is to identify renewable energy sources.
One such source could be a nuclear fusion reactor. Today’s nuclear reactors are fission based and produce radioactive waste. Fusion promises that it can deliver large amounts of energy, without the same waste as fission.
At ISC 2022, Nvidia is announcing that the UK Atomic Energy Authority (AEA) is using the Nvidia Omniverse Simulation Platform to accelerate the design and development of a full-scale fusion reactor.
“With the Nvidia Omniverse, researchers could potentially create a fully functional digital twin of the reactor, helping to ensure that the most efficient design for construction is selected,” Harris said.
The goal for Omniverse and the digital twin is to have an AI-generated replica of the fusion reactor system. The UK AEA also plans to emulate the physics of fusion plasma containment., The simulation will be done with the Nvidia modulus AI-Physics framework, to actually model the fusion reaction and how it can be controlled.
“The sacred grail of fusion energy is not only capable of creating a fusion reaction, but it is sustainable,” Harris said. “We really think this will be the path to sustainable energy.”