Scientific researchers need massive computing resources that can support research wherever it takes place. Whether they’re conducting groundbreaking pharmaceutical research, exploring alternative energy sources, or discovering new ways to prevent financial fraud, the state-of-the-art AI computing resources available are key to driving innovation. This new computing model could solve the challenges of generative artificial intelligence and usher in the next wave of innovation.
Cambridge-1, an NVIDIA supercomputer launched in the UK during the pandemic, has been the source of discoveries by some of the country’s leading health researchers. The system now becomes a part NVIDIA DGX Cloud to accelerate scientific innovation and discovery—in nearly every field.
As a cloud-based resource, it will expand access to AI supercomputers for researchers in climate science, autonomous machines, worker safety and other fields, delivered with the simplicity and speed of the cloud, ideally located for access in the UK and Europe.
DGX Cloud is a multi-node AI learning service that enables any enterprise to access advanced supercomputing resources from a browser. The initial Cambridge-1 infrastructure included 80 NVIDIA DGX systems; it will now join forces with DGX Cloud to provide customers with access to world-class infrastructure.
History of Healthcare Insights
Academia, start-ups and the UK’s large pharma ecosystem have used the Cambridge-1 supercomputing resource to accelerate research and develop new approaches to drug discovery, genomics and medical imaging using generative artificial intelligence in one of the following ways:
- InstaDeep, in collaboration with NVIDIA and the Technical University of Munich lab, developed a 2.5 billion parameter LLM for genomics at Cambridge-1. This project aimed to create a more accurate model for predicting the properties of DNA sequences.
- King’s College London used to create Cambridge-1 100,000 synthetic brain images — and made them available free of charge to health researchers. Using an open source AI imaging platform MONAIresearchers at King’s have created realistic, high-resolution 3D images of the human brain by training for weeks, not months.
- Oxford Nanopore used Cambridge-1 to rapidly develop highly accurate, efficient models for base identification in DNA sequencing. The company also used the supercomputer to support findings for the ORG.one project, which aims to sequence the DNA of endangered species.
- Peptone, in collaboration with a pharmaceutical partner, used Cambridge-1 to run physics-based simulations to assess the effect of mutations on protein dynamics to better understand why specific antibodies work effectively. This research could improve antibody development and biologic drug discovery.
- Relation Therapeutics has developed a large language model that reads DNA to better understand genes, a key step toward developing new drugs. Their research brings us closer to understanding how genes are controlled in certain diseases.