Nvidia and Dassault Systemes are advancing the next frontier of artificial intelligence by developing what they call "industry world models" designed to power physical AI across industries. At the heart of this initiative is the belief that the true potential of AI lies not just in digital simulations, but in its ability to understand and interact with the physical world through accurate, physics-based modeling. Rev Lebaredian, vice president of Omniverse and simulation at Nvidia, emphasized that the future of AI is physical AI—systems grounded in real-world laws of physics, capable of reasoning about cause and effect, imagining possible futures, and making intelligent decisions. To achieve this, the world must first be accurately modeled in digital environments. This is where the expanded 25-year partnership between Nvidia and Dassault comes in. Dassault, a pioneer in digital twin technology for over 45 years, is combining its deep industrial expertise and decades of engineering knowledge with Nvidia’s AI infrastructure, open models, and software libraries. The goal is to create a shared platform that enables scalable, intelligent, and science-validated simulation across fields such as engineering, materials science, digital biology, and manufacturing. These industry world models go beyond general AI by embedding first principles—such as physics, engineering laws, and system constraints—alongside four decades of real-world industrial experience. They integrate multi-scale, multi-discipline modeling with AI, enabling systems to reason, predict, and act with intent. As Florence Hu-Aubigny, executive vice president of R&D at Dassault, explained, these models ensure outputs are not only innovative but also aligned with real-world industrial realities. The partnership was announced at Dassault’s 3DExperience World conference in Houston. It builds on Dassault’s 3D Universes, a digital environment that combines generative AI, real-time IoT data, and advanced simulation. Within this ecosystem, AI-powered virtual companions—Aura for business, Leo for engineering, and Marie for scientific research—act as intelligent assistants that understand user intent, reason using industry world models, and automate complex workflows. To scale this vision, Dassault is establishing AI factories powered by Nvidia’s infrastructure, available through its Outscale cloud platform across three continents. These AI factories will run on Dassault’s 3DExperience platform, ensuring data sovereignty, security, and protection of intellectual property. Nvidia, in turn, is using Dassault’s model-based systems engineering (MBSE) technology to design its own AI factories, starting with the Rubin platform. This technology will also be integrated into Nvidia’s Omniverse DSX Blueprint, an open reference design for building large-scale AI data centers. Key use cases include the integration of Nvidia’s BioNeMo platform with Dassault’s Biovia world models to accelerate drug and material discovery. Dassault’s Simulia virtual twins will leverage Nvidia’s CUDA-X and AI physics libraries for faster, more accurate engineering predictions. Nvidia’s Omniverse physical AI libraries will enhance Delmia’s virtual twins for global production systems, supporting autonomous and software-defined manufacturing. Additionally, Nvidia’s Nemotron open models and AI technologies will be embedded into the 3DExperience platform, empowering the next generation of agentic AI assistants. This collaboration reflects a broader shift in industry, where 56% of manufacturers are already piloting AI-driven smart manufacturing, and 95% plan to invest in AI, machine learning, or generative AI within five years. By combining Dassault’s industrial depth with Nvidia’s AI scale, the two companies are laying the foundation for a new era of intelligent, physics-grounded innovation.