AivexaNewsSearch
AI news for builders and product teamsChecked every hour

NVIDIA Announces New Investments in the UK Startup Ecosystem

Collected Oct 14, 2026

NVIDIA used a UK-focused announcement to disclose planned investments in five British startups and to lay out a broader picture of how its hardware, models and software are being adopted across British industry. The companies named as new investment targets are Boltz, Foresight, Hiverge, Prima Mente and Riverlane. The announcement did not disclose the size of any individual investment. NVIDIA said it has already been backing UK startups including Basecamp Research, Ineffable Intelligence, Synthesia and Wayve.

The startups span several technical domains. Boltz is a London-based AI company founded by MIT researchers that builds biomolecular models and AI agents intended to help scientists predict molecular structures and interactions and design potential new medicines. Foresight, also London-based, works with owners of major infrastructure projects, hyperscale data centers among them, on predicting completion dates, mitigating risk and shortening the path to revenue. Hiverge, from Cambridge, was founded by researchers from Google DeepMind and the University of Cambridge and is building what it describes as an AI discovery engine that writes and optimizes algorithms automatically for tasks such as faster AI training and complex logistics planning. Prima Mente, a London AI biology startup, is developing biological foundation models aimed at decoding complex diseases and identifying new therapeutic targets, beginning with Alzheimer's, Parkinson's and ALS. Riverlane, based in Cambridge, builds real-time quantum error correction hardware and software that it says work across all major qubit types, with the goal of making quantum computers reliable enough for practical use.

Separately, NVIDIA is joining the Earthshot Prize as a Global Alliance Founding Partner, contributing compute and technical expertise to innovators working on climate and sustainability problems. The company also reiterated a five-year, $250 million compute commitment aimed at strengthening agentic safety and security research in the UK.

The rest of the announcement is a deployment round-up rather than new corporate news, and it shows how NVIDIA's stack is being picked up across sectors. In energy, Fuse Energy is moving its customer-facing AI onto in-house infrastructure built on NVIDIA B300 Blackwell Ultra, which it says delivers faster and more responsive support to more than 450,000 UK households. Sovereign AI company Locai Labs released Nemotron-Personas-UK, an open source dataset of one million synthetic personas intended to reflect the UK's demographic, regional and cultural diversity for developers building more representative systems.

Healthcare accounts for the largest cluster. OneAdvanced says it is deploying what it calls the UK's first and most advanced fine-tuned LLM for primary care AI triage, built with NVIDIA Agent Toolkit software and open models, and serving millions of NHS patients. Heidi has integrated NVIDIA Nemotron open models so clinicians can reach trusted medical evidence faster. Speechmatics is deploying Oak, a multilingual medical speech recognition model, on NVIDIA Triton Inference, which gives healthcare organizations control over where sensitive patient audio is processed. Apian, Synnovis and NVIDIA are working together on advanced computing and digital simulations using NVIDIA Cosmos to let hospital robots move lab samples, freeing clinical staff for patient care. Danaher and NVIDIA are collaborating on autonomous science in laboratory workflows and instruments; Danaher's AI-powered autonomous lab, designed for up to eight times faster discovery of affinity reagents, uses NVIDIA BioNeMo open models inside its AI-driven protein design workflows.

On the civic side, Liverpool City Region is building an NVIDIA-powered "City Brain," framed as a sovereign AI strategy and a blueprint for dozens of unitary authorities across the UK. Digital Jersey, Innovation City Belfast and Tees Valley Combined Authority are each exploring how trusted AI compute can serve communities and public services.

Financial services is the other dense cluster. LSEG and NVIDIA are bringing GPU-accelerated market analytics to global institutions. Revolut's PRAGMA, a foundation model for banking event sequences pre-trained on 26 million anonymised user histories and introduced earlier in the year, has moved from research into real-time scaling. Lloyds Banking Group built ATUM, a transaction foundation model that turns billions of banking transactions into reusable AI, while Starling is developing a banking foundation model trained on Civo's UK sovereign cloud infrastructure. NatWest is collaborating with NVIDIA on Nemotron Voice to better understand how customers across the UK speak, and Modulr is deploying a network of agents using Nemotron across distinct parts of payment operations. For developers, the pattern worth noting is that several of these are domain-specific foundation models trained on proprietary institutional data rather than general-purpose chat models, with sovereignty of where data is processed treated as a first-class requirement.

Quantum computing rounds out the list. NVIDIA said Riverlane has demonstrated that GPUs can accelerate decoder recalibration workflows used in quantum error correction, helping decoder performance keep pace with real-time changes in quantum hardware behavior; the work sits inside a broader collaboration on combining AI, accelerated computing and quantum error correction. ORCA Computing unveiled PT-3, its latest photonic quantum accelerator, at the Digital Realty Innovation Lab, and ran the NVIDIA Proteina model for protein design on it, reporting that its hybrid quantum-classical extension improved identification of useful proteins by 20 percent.

NVIDIA also noted that more than 3,100 UK companies are part of its Inception program for startups. Jensen Huang, NVIDIA's founder and CEO, is scheduled to deliver a GTC Berlin keynote on Wednesday, Oct. 21, at 11 a.m. CEST.

Why it matters: For UK founders, the practical signal is access — capital plus compute, with Inception acting as an on-ramp for smaller teams. For product teams in healthcare, finance and the public sector, the more consequential items are the sovereign and regulated-data deployments, where where the inference runs is as much a design constraint as the model itself. That constraint is likely to shape vendor choices in the UK market more than raw model quality in the near term.

Read at NVIDIA Blog

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Across industries, the U.K. is putting the full stack to work. The scale of deployment is growing fast, from enterprises transforming their businesses to startups solving some of the planet’s hardest problems. NVIDIA is putting that principle into practice, including through a five-year, $250 million compute commitment to strengthen the U.K.’s agentic safety and security […]