NVIDIA Selects Yunar to Build First Human-like Robot AI Platform, Boosting Investment in 'Physics AI' Strategy

On June 1, during its Taipei GTC conference, NVIDIA announced it has tapped Chinese robotics firm Ubtech as its first humanoid robot AI platform partner for university and research institute sales. The platform will integrate Ubtech’s H2 humanoid robot with NVIDIA’s Jetson Thor computing platform, the Isaac GR00T humanoid robot AI model, and simulation software, with early research partners including Stanford University, ETH Zurich, and other such institutions. NVIDIA CEO Jensen Huang noted the platform is built to lower barriers to humanoid robot development by combining robot hardware, AI models, data generation tools, and simulation software to give researchers a plug-and-play development environment. He emphasized that after generative AI, “physical AI” will mark the next phase of artificial intelligence advancement. For reference, the Ubtech H2 robot stands roughly 1.8 meters tall, is equipped with NVIDIA’s Jetson Thor chip and Blackwell GPU, enabling AI models to run directly on the robot itself. The platform also includes a dexterous hand system from Singapore-based robot hand manufacturer Sharpa. NVIDIA added that an upgraded version, the H2 Plus robot, is set to launch this October and will be available for general public purchase. The first institutions adopting this platform include the Stanford Robotics Center, ETH Zurich, UC San Diego’s Advanced Robotics Lab, and U.S. AI research institute Ai2. Notably, this collaboration comes as Ubtech prepares to enter the capital markets. Public data shows Ubtech is planning an initial public offering (IPO) on the Shanghai Stock Exchange’s Sci-Tech Innovation Board, aiming to raise 4.2 billion yuan. The company also disclosed that over 40% of its revenue originates from markets outside China. Earlier, Huang has stated that robotics is poised to be one of NVIDIA’s largest growth opportunities over the next five years, and he projects the “physical AI” market could eventually reach trillions of dollars in value.