NVIDIA at VivaTech 2026

VivaTech 2026 takes place June 17–20 in Paris, France at the Paris Expo Porte de Versailles, which is located at 1 Place de la Porte de Versailles, 75015 Paris, France. It is Europe's largest startup and technology event, bringing together enterprises, investors, developers, and innovators from across the globe.

Registration and official event information are available at vivatechnology.com.

NVIDIA and its partners are showcasing how AI is moving from experimentation into real-world production across industries. Key areas of focus include agentic AI, physical AI and robotics, sovereign AI infrastructure, and open-source model development. Demonstrations and sessions span applications in pharma, manufacturing, automotive, retail, and financial services — built on the NVIDIA AI platform.

Sovereign AI refers to a region or entities' ability to develop, own, and operate its own AI capabilities — including the infrastructure, data, and models that power them. For Europe, this means building AI systems on locally controlled compute and purpose-built models trained on national languages and datasets. NVIDIA provides the full-stack platform — from GPU infrastructure to model development tools like NVIDIA NeMo and Nemotron — to enable governments, enterprises, and research institutions to build sovereign AI capabilities.

NVIDIA Nemotron is a family of open large language models (LLMs) optimized for enterprise and sovereign AI deployments. Nemotron models are designed for high performance on NVIDIA hardware and are available for customization, fine-tuning, and deployment through NVIDIA NIM microservices. At VivaTech 2026, Nemotron will feature prominently in discussions around open-source AI and European AI sovereignty.

Agentic AI refers to AI systems that can autonomously plan, reason, and execute multi-step tasks — going beyond simple question-and-answer interactions to taking action in the real world. NVIDIA enables agentic AI through its NIM microservices and the NVIDIA AI Enterprise software platform, which give developers the tools to build, deploy, and orchestrate AI agents at enterprise scale. Sessions at VivaTech 2026 will explore how agentic AI is being applied across industries including pharma, finance, and enterprise operations.

Physical AI describes intelligent systems that can perceive, reason, and act in the physical world — the foundation of modern robotics and autonomous machines. NVIDIA enables physical AI through three key platforms: NVIDIA Cosmos, a world foundation model that generates synthetic training data and simulation environments for robots and autonomous vehicles; NVIDIA Isaac, an end-to-end robotics development platform; and NVIDIA Omniverse, a platform for building physically accurate digital twins and simulation environments. Together, they form a complete stack for developing, training, and deploying physical AI systems.

NVIDIA Cosmos is a world foundation model platform that accelerates the development of physical AI. It generates photorealistic synthetic data and simulated environments to train and validate AI models for robotics and autonomous systems, dramatically reducing reliance on costly real-world data collection. Cosmos enables developers to move from simulation-based training to real-world deployment at scale — a central theme in the physical AI sessions at VivaTech 2026.

NVIDIA Omniverse is a platform for building and running physically accurate 3D simulations, digital twins, and virtual worlds. Enterprises use it to simulate factories, industrial processes, and products before physical deployment — reducing costs and accelerating development cycles. At VivaTech 2026, Omniverse features in demonstrations of industrial AI and robotics, including joint work with Microsoft and enterprise customers on real-world digital twin applications.

NVIDIA and its partners are presenting sessions across the main program and partner booths. Topics include scaling AI agents in the pharmaceutical industry (featuring Sanofi and Biolevate), intelligent mobility and automotive AI (featuring Stellantis and Capgemini), AI-powered applications for e-commerce (featuring Nebius and Photoroom), and a fireside panel on open-source models and sovereign AI. Our partners are hosting theater sessions on their booths, covering agentic AI with Nemotron, physical AI and robotics, and enterprise AI deployment. View sessions at vivatechnology.com/sessions.

NVIDIA Inception is a free program that supports AI startups with access to the NVIDIA platform, technology, and ecosystem. More than 700 Inception member companies are active in France alone. At VivaTech 2026, Inception startups are showcased across the show floor — including a dedicated startup village within the AWS booth and a joint NVIDIA-Microsoft GenAI Studio area featuring Inception companies building AI applications on Microsoft Azure. VivaTech is one of the key venues where Inception startups connect with enterprise customers, investors, and partners.

Learn more about NVIDIA Inception.

NVIDIA's Deep Learning Institute (DLI) offers hands-on, self-paced training in AI, data science, robotics, and accelerated computing. Courses range from introductory to expert level, with certification available. Many courses are available free or at a discount through the NVIDIA Developer Program — which also provides access to tools, SDKs, early software releases, and a global developer community.

Explore DLI Training and join the Developer Program.

NVIDIA On-Demand hosts recordings from GTC and other NVIDIA events, including technical sessions, keynotes, and deep dives on topics like sovereign AI, agentic applications, physical AI, and open-source model development. It's the best place to explore AI content beyond what's covered live at VivaTech.

Browse NVIDIA On-Demand.