NVIDIA and Dassault Systèmes Unite to Build Next-Generation Industrial AI Infrastructure

2026-02-04 companies

Santa Clara, Wednesday, 4 February 2026.
Announced February 3, this alliance integrates Dassault’s virtual twins with NVIDIA’s computing to create “Physical AI,” a groundbreaking system anchored in the laws of physics rather than simple prediction.

Defining Physical AI

Formalized on February 3, 2026, this strategic partnership represents a significant pivot from generative AI to “physical AI,” a concept NVIDIA CEO Jensen Huang describes as the next frontier of artificial intelligence [1][3]. Unlike traditional models that primarily predict or generate text and images, this new platform utilizes the laws of physics and validated scientific data to accurately understand and simulate the real world [1][7]. Pascal Daloz, CEO of Dassault Systèmes, emphasized that this collaboration marks an era where AI becomes a “force multiplier for human ingenuity,” moving beyond mere generation to actual comprehension of physical constraints and predictive reliability [1].

The Four Pillars of Industrial Transformation

The collaboration is structured around four key sectors: biology, engineering, factories, and the workforce [1]. In the biological sphere, the companies are combining NVIDIA’s BioNeMo with Dassault’s BIOVIA to model complex biological systems [1]. For engineering and manufacturing, the integration involves connecting NVIDIA’s Omniverse and CUDA-X libraries with Dassault’s SIMULIA and DELMIA applications to create highly accurate digital twins [1]. Additionally, the partnership introduces “Virtual Companions” designed to support industrial jobs, leveraging the massive datasets inherent in industrial operations to assist researchers and engineers [1].

Real-World Applications: From EVs to CPG

Major industry players are already leveraging these technologies to accelerate development cycles. Lucid Motors is utilizing the platform to transition from concept to production “faster than ever,” while the French food giant Bel Group is modeling and optimizing its products on a large scale [1][3]. In the academic and research sector, the National Institute for Aviation Research (NIAR) in the U.S., the largest university aviation R&D institution in the country, has adopted these “Virtual Companions” to advance aviation research [1][3]. These implementations highlight the shift toward the industrial metaverse, where virtual twins reshape decision-making across industries [2].

Infrastructure Sovereignty and Rubin Architecture

Addressing critical concerns regarding data privacy and industrial confidentiality, Dassault Systèmes is deploying “AI factories” through its Outscale cloud brand [1]. These facilities will span three continents to ensure strict data sovereignty, compliance, and intellectual property protection for clients [1]. Conversely, NVIDIA is adopting Dassault’s Model-Based Systems Engineering (MBSE) capabilities to design its own future infrastructure, specifically the Rubin platform [1]. The Rubin architecture, which features the Vera CPU and Rubin GPU, is slated for production in the second half of 2026 and boasts bandwidth capabilities of 3.6TB/s per GPU [5].

Market Implications

Following the announcement, NVIDIA (NVDA) continues to assert its dominance in the accelerated computing sector. As of February 3, 2026, NVIDIA’s stock price stood at $180.34, with a market capitalization of approximately $4.51 trillion [5]. This valuation reflects the market’s confidence in NVIDIA’s expanding ecosystem, which now includes this critical bridge to the industrial metaverse. Investors will likely look for further details when the company discusses its fiscal year 2026 results in a conference call scheduled for February 25, 2026 [5].

Sources

Industrial AI Digital Twins