NVIDIA backs French and UK AI startups with new compute investments
NVIDIA announces funding for agentic computing in France and the UK, supporting new server production and satellite launches.

Key points
- NVIDIA commits $250 million to UK agentic safety and startup infrastructure over five years.
- Bull will produce NVIDIA Vera Rubin Ultra NVL72 servers in its Angers, France factory by 2028.
- French companies are deploying NVIDIA Nemotron models for sovereign public sector and financial services.
NVIDIA is increasing its investment in artificial intelligence infrastructure for startups in France and the United Kingdom, according to blog posts published by the company. The announcements detail new hardware production plans in Europe and financial commitments to support agentic computing ecosystems in both nations.
In plain English
Agentic computing refers to software that can perform tasks autonomously rather than just answering questions. Unlike simple chatbots, these agents can plan steps, use tools and execute complex workflows. This technology allows companies to automate processes that previously required human oversight.
The background
In France, the ecosystem has moved from planning to application, supported by the AI Action Summit in February 2025 and Choose France Summits in 2025 and 2026. Bull is extending its European AI server production to include NVIDIA Vera Rubin Ultra NVL72 systems, with integration at its Angers factory inaugurated on Oct. 2.
Altair-1, the first satellite in a $1 billion, 50-satellite constellation, launched this month. The satellite, integrated by Loft Orbital and built by Orbitworks, uses NVIDIA Jetson AGX Orin modules and Mistral models for autonomous operations. It successfully reached orbit to conduct real-time Earth observation.
NVIDIA Nemotron open models are now in production for French organizations requiring sovereign data handling. LINAGORA and Docaposte secured the CANUT framework agreement, making Twake.ai the referenced sovereign digital workplace for over 2,500 public organizations. In finance, SBS is launching a platform for banks to deploy governed agentic services.
Mistral introduced the Mistral Large 4 model under private preview last week. The open weight model targets cyber defense, manufacturing and finance. Fondation Inria, with NVIDIA support, is launching a working group on open source modeling, with findings expected in spring next year.
L’Oréal is using the NVIDIA ALCHEMI Toolkit for in-silico cosmetic screening, aiming for discovery up to 100 times faster. Mirakl onboarded 70 million products using NVIDIA GPUs, cutting catalog onboarding to under two hours. Photoroom and Vestiaire Collective are applying visual modeling to marketplace performance, while Omi uses 3D digital twins for faster inference.
ThinkDeep and Akila built a digital-twin agent for Côte-d’Or, managing nearly 5,800 kilometers of infrastructure. The system, powered by NVIDIA Cosmos and Nemotron, cut operating costs by 10% with targets of 30% reduction. Port Sud de France uses similar technology for goods processing and security.
Enchanted Tools is deploying social robots in Paris for pediatric cancer care. In telecom, Bouygues Telecom and Prisme.ai launched AI Studio V2, powered by NVIDIA infrastructure, allowing employees to build agents securely. Across retail, public services and science, open models and NVIDIA infrastructure are driving consistent results.
What changes now
The United Kingdom is receiving a five-year, $250 million compute commitment from NVIDIA to strengthen its agentic safety and startup infrastructure. This investment supports enterprises and startups solving complex problems across industries. The scale of deployment in the UK is growing rapidly, putting the full stack to work for business transformation and global challenges.
Key terms explained
Agentic computing involves AI systems that act independently to achieve goals. Teams usually define clear boundaries and safety protocols for these agents. Check whether your organization has policies for autonomous decision-making. Verify that agents operate within approved parameters and do not access sensitive data without authorization. Monitor agent actions to ensure alignment with business objectives and regulatory requirements.
What to do and how to stay safe: NVIDIA
- Review current AI agent deployments to ensure they operate within defined safety boundaries.
- Verify that sovereign data requirements are met when using public sector AI platforms.
- Monitor infrastructure updates for compatibility with new NVIDIA hardware releases.
- Assess open model integrations for security compliance in financial and public services.
Step-by-step guide: Stop AI Chatbots Leaking Sensitive Business Data
General security guidance from the Vector Update newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
What is NVIDIA investing in France?
NVIDIA is supporting agentic computing infrastructure, including Bull’s server production and various startup deployments across industries.
How much is NVIDIA investing in the UK?
NVIDIA announced a five-year, $250 million compute commitment to strengthen the UK’s agentic safety and startup ecosystem.
What is Altair-1?
Altair-1 is the first satellite in a 50-satellite constellation, using NVIDIA modules and Mistral models for autonomous Earth observation.



