NVIDIA is signaling that the era of robots moving from controlled demos to active deployment on factory floors has arrived.
At a recent event in Seoul, NVIDIA CEO Jensen Huang declared that 'physical AI'—intelligent robots that can perceive and interact with the physical world—is becoming a reality. This isn't just a futuristic vision; it's a concrete strategy backed by deepening partnerships with Korean industrial leaders like Hyundai Motor Group and LG. The goal is to integrate AI-driven robotics directly into manufacturing, transforming how things are made.
This acceleration is built on a clear, multi-year foundation. First, the groundwork was laid through strategic alliances. A key moment was the March 2026 announcement of an 'AI factory' collaboration between NVIDIA and Hyundai. This pact established a formal program to use NVIDIA's entire technology stack, from chips to software, to build next-generation smart factories.
Second, the core technology is now mature enough for real-world use. NVIDIA has developed crucial building blocks like the Jetson Thor, a powerful and compact computer designed to be the 'brain' for robots, and Project GR00T, a reference platform for creating humanoid robots. These tools are combined with NVIDIA's Omniverse, a simulation platform that allows companies to build a 'digital twin' of their factory. In this virtual world, they can safely train and test robots millions of times before deploying them alongside human workers, a critical concept known as sim-to-real.
Finally, the market itself is ready. NVIDIA's latest financial results revealed that its industrial AI division is now generating revenue on par with its traditional cloud business. This proves that companies are no longer just experimenting; they are making significant investments in AI for production. Hyundai, with its ownership of Boston Dynamics and its stated goal of deploying humanoid robots in its U.S. plant by 2028, is a perfect example of an anchor customer ready to lead this transition. Together, these factors signal that the industrial robotics revolution is moving from theory to practice.
- Physical AI: AI systems, typically robots, that can perceive, reason, and interact with the physical world, not just digital data.
- Digital Twin: A virtual replica of a physical object, process, or system. It's used for simulation, testing, and optimization before implementing changes in the real world.
- Sim-to-Real: The process of training an AI model in a simulated environment and then transferring that learned knowledge to a real-world robot or system.