From Earth to Orbit, France’s Ecosystem Delivers Agentic Computing at Scale With NVIDIA

Companies in France are deploying agentic computing across various sectors, including space, financial services, the public sector, and scientific research. This shift illustrates how the country’s ecosystem has transitioned from ambition to practical application.

France’s advancements are built on a foundation of strategic groundwork. The AI Action Summit in February 2025 and the Choose France Summits in 2025 and 2026 have anchored significant compute commitments within the country, leading to the development of a comprehensive ecosystem.

Sovereign AI Infrastructure

French companies are expanding their agentic infrastructure utilizing NVIDIA technology. Bull is extending its European AI server production to include NVIDIA Vera Rubin Ultra NVL72 systems, which are expected to be operational by 2028. The production of trays will occur in Europe, with rack integration taking place at its factory in Angers, inaugurated on October 2.

Additionally, Altair-1, the first satellite in a $1 billion, 50-satellite constellation featuring on-board autonomous capabilities, successfully launched this month and reached orbit. This satellite, integrated by Loft Orbital and constructed in Abu Dhabi by Orbitworks—a joint venture between Loft Orbital and Marlan Space—utilizes NVIDIA Jetson AGX Orin edge modules and runs Mistral’s models for real-time Earth observation and autonomous operations.

Open Models

NVIDIA Nemotron open models are being implemented across French organizations that prioritize sovereign and trustworthy data. LINAGORA and Docaposte have secured the CANUT framework agreement, making Twake.ai, built on NVIDIA Nemotron, the exclusive sovereign digital workplace for over 2,500 French public organizations.

In the financial services sector, SBS is launching a sovereign platform that enables banks to deploy governed generative and agentic services within their environments, in collaboration with NVIDIA. Last week, Mistral introduced the Mistral Large 4 open model under private preview, showcasing advanced performance in cyber defense, manufacturing, finance, and multimodal tasks, providing a robust foundation for model customization across key industries.

Supported by NVIDIA, Fondation Inria is initiating a working group focused on the adoption of open-source advanced modeling and tools, with findings expected in the spring of next year.

Agents at Work Across Industries

In various sectors such as retail, public services, telecommunications, and science, the trend is clear: open models and NVIDIA infrastructure are yielding significant results in France. L’Oréal is integrating NVIDIA ALCHEMI into its research workflows, allowing scientists to explore formulation possibilities in silico, which can accelerate the discovery process by up to 100 times compared to traditional methods.

Mirakl has successfully onboarded over 70 million products using fine-tuned open models on NVIDIA GPUs, reducing seller catalog onboarding to under two hours. Companies like Photoroom and Vestiaire Collective are also demonstrating the tangible business impact of advanced visual modeling on marketplace performance.

Omi assists brands in accurately representing products in AI-generated marketing visuals at scale, combining its approved 3D digital twins with open models optimized on NVIDIA for four times faster inference. In Côte-d’Or, ThinkDeep and Akila have developed a digital-twin agent, powered by NVIDIA Cosmos open world models and Nemotron, managing nearly 5,800 kilometers of roads and buildings, achieving a 10% reduction in operating costs with a target of 30% as deployment progresses.

Port Sud de France utilizes ThinkDeep’s DeepBrain platform, powered by Nemotron and Cosmos, for goods processing and port security. In the telecommunications sector, Bouygues Telecom and Prisme.ai have launched AI Studio V2, leveraging NVIDIA’s full-stack infrastructure to enable employees to build their own agents securely.

In quantum computing, Alice & Bob has integrated the NVIDIA NVQLink platform for its superconducting cat qubits, demonstrating how GPUs can enhance quantum error correction efficiency. Similarly, Quandela has utilized NVQLink to create an integration between a DGX Spark system and its photonic qubits, advancing QPU-GPU supercomputing and exploring hybrid quantum-classical applications.

Tune in to NVIDIA founder and CEO Jensen Huang’s GTC Berlin keynote on Wednesday, October 21, at 11 a.m. CEST.

Source: blogs.nvidia.com