NVIDIA Forms Strategic Alliances to Scale Physical AI Robotics

NVIDIA Corporation launched a series of strategic partnerships on the eve of its GPU Technology Conference in California to accelerate the global adoption of physical AI. The company collaborated with European semiconductor manufacturers Infineon, NXP Semiconductors, and STMicroelectronics to integrate sensors, motion control systems, and the Holoscan Sensor Bridge into hardware portfolios, reducing latency between robot sensors and processing units. To solve infrastructure fragmentation, NVIDIA also partnered with AI cloud provider Nebius to integrate the Physical AI Data Factory Blueprint into a global managed runtime for robotics developers.

Simultaneously, NVIDIA backed Skild AI, a startup that developed the Skild Brain foundation model. This model uses omni-bodied intelligence learned from human videos and simulations to eliminate task-by-task human programming. In a practical application of this technology, Skild AI deployed its model to Foxconn assembly lines in Houston, Texas, where robots now handle high-precision assembly of NVIDIA Blackwell GPU server racks.

To scale this intelligence beyond a single site, Skild AI partnered with ABB Robotics and Universal Robots to embed its software across wide industrial portfolios. This effort is supported by a $1.4 billion funding round for Skild AI led by NVIDIA and SoftBank, alongside SoftBank's $5.38 billion pending acquisition of ABB's robotics business. These initiatives collectively support NVIDIA's goal of producing AI supercomputers entirely within the United States through advanced domestic automation.