Nvidia on 14 October outlined new UK startup investments and AI deployments across energy, healthcare, finance and local government, positioning the country as a test bed for its technology stack. NVIDIA Newsroom
The company's announcement indicates UK institutions are moving from pilot projects to live systems built on Nvidia hardware, software and open models, alongside a parallel $250 million UK compute commitment for academic research on AI safety and security. NVIDIA Newsroom nvidianews.nvidia.com
Startup investments and climate partnership
Nvidia said it has backed several UK startups in recent years and has now "announced new planned investments" in five more: Boltz, Foresight, Hiverge, Prima Mente and Riverlane. NVIDIA Newsroom
According to Nvidia, Boltz, based in London, builds biomolecular models and AI agents for predicting molecular structures and interactions. Foresight, also in London, helps owners of large infrastructure projects predict completion dates and manage risk. Hiverge, a Cambridge startup founded by researchers from Google DeepMind and the University of Cambridge, builds an "AI discovery engine" for tasks such as faster AI training and logistics planning. Prima Mente, another London firm, focuses on biological foundation models for decoding complex diseases, initially in Alzheimer's, Parkinson's and ALS. Riverlane, based in Cambridge, develops real-time quantum error correction hardware and software. NVIDIA Newsroom
Nvidia said it is partnering with The Earthshot Prize as a "Global Alliance Founding Partner" to support climate and sustainability innovators. NVIDIA Newsroom The Earthshot Prize stated that Nvidia will provide finalists with technical training, engineering support and access to AI computing resources, including its Inception startup programme, educational material on open models and climate tools, and Nvidia DGX Spark systems. earthshotprize.org
Sector deployments from energy to healthcare
Nvidia's UK post lists deployments across regulated sectors. In energy, it said supplier Fuse Energy is moving its customer AI to in-house infrastructure built on its B300 Blackwell Ultra platform, to support more than 450,000 UK households. NVIDIA Newsroom
In healthcare, software provider OneAdvanced is deploying what Nvidia calls "the U.K.'s first and most advanced fine-tuned LLM for Primary Care AI-triage" using Nvidia Agent Toolkit and open models, with the stated goal of supporting millions of National Health Service patients. The announcement does not provide deployment timelines or independent validation of the model's performance. Heidi has integrated Nvidia Nemotron open models to provide clinicians with faster access to medical evidence, and Speechmatics is deploying Oak, a multilingual medical speech recognition model, using Nvidia's Triton Inference software for healthcare organisations to retain control over where patient audio is processed. NVIDIA Newsroom
Nvidia said it is teaming with Apian and Synnovis to use its Cosmos platform for simulating hospital robot lab-sample movement, and with Danaher to incorporate accelerated computing, "scientific and physical AI" and Nvidia BioNeMo open models into laboratory workflows. Danaher's "AI-powered autonomous lab" is designed to deliver "up to 8x faster discovery of affinity reagents". Nvidia's post does not include third-party metrics on these claimed accelerations. NVIDIA Newsroom
Smart city and financial deployments
Nvidia said the Liverpool City Region is building an "NVIDIA-powered 'City Brain'" as part of a sovereign AI strategy intended as a blueprint for dozens of UK unitary authorities. Digital Jersey, Innovation City Belfast and the Tees Valley Combined Authority are each exploring how "trusted AI compute" can support public services, without project details or timeframes offered. NVIDIA Newsroom
In financial services, Nvidia said it is working with LSEG to bring GPU-accelerated market analytics to global institutions and that multiple UK-based banks and payment firms are building foundation models on its stack. Revolut has introduced PRAGMA, a foundation model for banking event sequences pre-trained on 26 million anonymised user histories, now moved "from research to real-time scaling". Lloyds Banking Group has built ATUM, a transaction foundation model for converting billions of banking transactions into reusable AI features. Starling is developing a banking foundation model trained on Civo's UK sovereign cloud infrastructure. NatWest is collaborating on Nemotron Voice to understand UK banking customer speech, and payments firm Modulr is deploying a network of agents using Nemotron for payment operations. NVIDIA Newsroom
Compute and quantum collaboration
Nvidia's ecosystem blog links its UK deployments to a previously announced five-year, $250 million in-kind compute donation for UK "agentic safety and security" research. In a separate release, the company said the commitment will provide $50 million in compute access annually, primarily to the wider UK AI safety and security research community, with a portion for the British Open-Ended Learning and Discovery Lab (BOLD), the Science of Fundamental AI Research Lab (SOFAIR) and Imperial College London. nvidianews.nvidia.com
Compute will be delivered through UK-based Nvidia cloud partners with infrastructure hosted in the UK, with provider selection, technical architecture and commercial terms still to be agreed. The programme is intended to fund work on secure agents, open-model security, real-world threat mitigations, secure sandboxes and safety assessment benchmarks. nvidianews.nvidia.com
The blog frames these activities as part of a community of more than 3,100 UK companies in Nvidia's Inception startup programme, and points to Nvidia founder Jensen Huang's planned GTC Berlin keynote on 21 October, without specifying further UK policy or regulatory commitments. NVIDIA Newsroom