NVIDIA Accelerates Investment in France’s Startups, Powering a Thriving AI Ecosystem

Companies in France are increasingly adopting agentic computing across various sectors, including space, financial services, and scientific research. This shift from ambition to tangible application showcases the country's vibrant tech ecosystem.

Structured initiatives have laid the groundwork for France's robust AI landscape. The AI Action Summit in February 2025 and the subsequent Choose France Summits in 2025 and 2026 marked significant commitments to computing power within the country, reinforcing a full-stack approach to AI development.

NVIDIA’s Investment Strategy in France

NVIDIA’s investment in France's startup ecosystem is pivotal, having previously backed innovative firms like Alice & Bob, Gradium, and Scintil Photonics. Across Europe, NVIDIA's total investment commitment approaches $10 billion.

Recently, NVIDIA announced plans to invest in a range of French startups, including Doctolib, Entalpic, Genesis AI, and Pigment. Each of these companies is tackling pressing challenges in their respective fields.

For instance, Doctolib is enhancing healthcare efficiency by developing digital agents that ease the workload of healthcare professionals, ultimately improving patient care access and promoting preventive medicine.

Entalpic, based in Paris, focuses on advanced material engineering at the atomic level, specializing in surfaces and interfaces crucial for semiconductor manufacturing and next-gen electronics.

Genesis AI is creating versatile robots outfitted with advanced intelligence capabilities through foundational models, robust data collection systems, and a practical robotics infrastructure.

Pigment develops a business planning platform, enabling organizations to enhance their financial forecasting, scenario modeling, and operational decision-making processes.

Establishing Sovereign AI Infrastructure

Companies in France are actively expanding their AI infrastructure using NVIDIA technology. For instance, Bull is extending its AI server production capabilities with the upcoming NVIDIA Vera Rubin Ultra NVL72 systems, expected in 2028, with production facilities fully based in Europe.

Additionally, the first satellite of a $1 billion, 50-satellite constellation, Altair-1, was launched this month, equipped with on-board autonomous systems. This satellite, developed by Loft Orbital in collaboration with Orbitworks, utilizes NVIDIA Jetson AGX Orin modules to enable real-time Earth observation and autonomous operations in orbit.

Leveraging Open Models

NVIDIA’s Nemotron open models are being integrated across various French organizations that prioritize sovereign and reliable data. For example, LINAGORA and Docaposte have established a framework agreement under CANUT, appointing Twake.ai—built on Nemotron—as the exclusive digital workspace for over 2,500 French public entities.

In the banking sector, SBS is introducing a sovereign AI platform with NVIDIA's collaboration, allowing banks to deploy regulated generative AI services within their infrastructures.

Dassault Systèmes is leveraging Nemotron and accelerated computing to advance its Industrial AI framework, speeding up the training of its Industry World Models and enhancing its agentic Virtual Companions platform grounded in industrial applications.

Mistral recently unveiled the Mistral Large 4 open model—under private preview—with capabilities in cyber defense and mission-critical industries, paving the way for customizable models across sectors. Additionally, Fondation Inria, with NVIDIA’s backing, is initiating a working group to promote open-source modeling and tool adoption.

AI Agents Across Diverse Industries

The continuing trend of open models and NVIDIA infrastructure is fostering AI-driven innovations across diverse fields such as retail, public services, telecommunications, and scientific research.

L’Oréal has incorporated NVIDIA ALCHEMI into its research frameworks, facilitating faster exploration of material formulations and guiding lab research significantly quicker than conventional methods.

Furthermore, Mirakl has streamlined product onboarding to under two hours, processing over 70 million items using optimized open models on NVIDIA’s infrastructure.

Companies like Photoroom and Vestiaire Collective illustrate the tangible marketplace impacts driven by advanced visual modeling.

In Côte-d’Or, ThinkDeep and Akila have developed a digital twin agent utilizing NVIDIA Cosmos open world models, managing extensive infrastructure and reducing operational costs by a projected 30% as deployment advances.

Telecommunications firms like Bouygues Telecom are utilizing NVIDIA’s full-stack infrastructure to empower employees to develop their AI agents in a safe environment.

In quantum computing, Alice & Bob has integrated the NVIDIA NVQLink platform into their superconducting systems, enhancing quantum error correction capabilities. Similarly, Quandela is employing NVQLink to facilitate hybrid quantum-classical applications.

For more insights, tune in to NVIDIA founder and CEO Jensen Huang’s GTC Berlin keynote on Wednesday, Oct. 21, at 11 a.m. CEST.