Nvidia has unveiled the culmination of a $100 million investment in supercomputing in the form of the Cambridge-1, which the company described as the United Kingdom’s fastest, most powerful supercomputer.
The Cambridge-1 is built on an Nvidia DGX SuperPOD supercomputing cluster involving 80 DGX A100 nodes, 640 A100 GPUs to deliver 400 petaflops of AI compute. These elements help Cambridge-1 attain the No. 41 spot on the list of the top 500 supercomputers worldwide.
The DGX SuperPOD architecture was key in enabling Nvidia “to build a supercomputer in record time, even during a pandemic and the restrictions that presents,” said David Hogan vice president, enterprise, EMEA at Nvidia. “Normally it would take a supercomputer several years to be built.”
Hogan said Nvidia built the new supercomputer in the U.K. because the country has become an important center of the life sciences industry, but also because the country has invested aggressively in the concept of using AI technology at the point of care. “It’s a place where Nvidia can bring its expertise in collaboration with some world-leading sciences to be transformative,” he said.
The Cambridge-1 leverages the Nvidia Clara AI application framework designed for the medical, health and life sciences industries. Its AI and simulation capabilities will be used to delve deeper into pressing challenges in those sectors, Nvidia said. Notably, the Cambridge-1 also is powered completely by renewable energy.
Its initial projects are with some of the biggest names in the health and sciences arenas, including AstraZeneca, GSK, Guy’s and St Thomas’ NHS Foundation Trust, King’s College London and Oxford Nanopore Technologies. For example, Nvidia is teaming with AstraZeneca to fuel faster drug discoveries by creating a transformer-based generative AI model for chemical structures. Such transformer-based neural network architectures are a fairly recent development, and enable researchers to leverage massive datasets using self-supervised training methods, avoiding the need for manually labeled examples during pre-training.
Meanwhile, King’s College London and Guy’s and St Thomas’ NHS Foundation Trust are using Cambridge-1 to teach AI models to generate synthetic brain images by learning from tens of thousands of MRI brain scans, from various ages and diseases. The ultimate goal is to use this synthetic data model to gain a better understanding of diseases like dementia, stroke, brain cancer and multiple sclerosis and enable earlier diagnosis and treatment.
According to a report by Frontier Economics, an economics consulting firm, Cambridge-1 has the potential to create an estimated value of £600 million (about $825 million) over the next 10 years.
RELATED: Nvidia CEO Jensen Huang and the future of… everything