(Yicai) Feb. 4 -- Dassault Systèmes is linking arms with US chip giant Nvidia to develop an industrial artificial intelligence platform that supports Virtual Twin technology, which provides digital replicas of real-world manufacturing environments that enable real-time simulation and optimization, the French software company said yesterday.

The long-term strategic partnership will combine Dassault’s Virtual Twin technology with Santa Clara-based Nvidia’s AI infrastructure, open models and accelerated software libraries to build scientifically validated industry World Models and empower designers and other professionals to acquire new knowledge through intelligent virtual assistants running on Dassault’s 3DEXPERIENCE platform.

“Together with Nvidia, we are building industry World Models that unite Virtual Twins and accelerated computing to help professionals across industries design, simulate and operate complex systems in areas such as biology, materials science, engineering and manufacturing,” said Pascal Daloz, chief executive officer of the Vélizy-Villacoublay-based firm.

The tie-up will also lay a new foundation for industrial AI, which could unlock greater innovation at scale in the emerging generative economy, he added.

Dassault unveiled three AI assistants built on industrial-grade World Models at the company’s annual summit for the user community of its SOLIDWORKS and 3DEXPERIENCE platforms on Feb. 2, that are aimed at accelerating the adoption of AI in industrial applications.

Companies that deploy these AI technologies could boost productivity by more than 10-fold, Daloz said.

Nvidia CEO Jensen Huang, who was at the Dassault event yesterday, said physical AI represents the next major frontier for AI. This is a massive market with a potential value of up to USD9 trillion, he said.

Huang told Yicai that the coming decade will be a period of explosive growth for robotics and profound changes will occur across the entire industry.

"Taking factories as an example, we will see robot systems of all sizes, including humanoid robots, emerge over the next decade," Huang said. "These robots can be called general-purpose or intelligent robots and will even be capable of reasoning, such as knowing that they need to open a door first before retrieving an object."

Since robots are programmed to perform specific tasks, they are only valuable when used to carry out a large number of repetitive and simple tasks and are used on a large scale, Huang said. Many companies need smarter and more intelligent robots. With the industrial AI platform built by Dassault and Nvidia, robots can use AI-driven automation to learn about the world and understand how to operate within it, just like humans and animals.

“This will allow companies of all sizes and across all industries to use AI to enhance their value, rather than being left behind by it,” Huang said.

As AI tools become more widely adopted, Dassault’s business model will change. AI opens the door to entirely new pricing models, Daloz said. While the company has charged for software licensing for decades, it may shift toward usage-based pricing as AI continues to advance.

Editor: Kim Taylor