In this collaboration, NVIDIA leads the core "physical AI" technology, while Foxconn, leveraging its manufacturing advantages, is responsible for mass production. As a long-term partner in NVIDIA's AI ecosystem, Foxconn has begun preparing production lines. Its US factories will simultaneously produce the humanoid robot and the Blackwell Ultra GB300 AI server – the core hardware supporting AI computing power. Their collaborative production provides the underlying computing power for the robot's "AI brain."
NVIDIA defines "physical AI" as the next generation of key areas following AI training and generative AI. It has been revealed that this robot will be equipped with a large language model and a dedicated "AI brain," aiming to transcend the scenario limitations of traditional robots and achieve autonomous perception and interaction with the physical world. Nvidia CEO Jensen Huang has previously emphasized that the impact of "physical AI" will far exceed that of generative AI, as it can penetrate deep into physical fields like commerce and industry, directly transforming daily life.
With its November debut approaching, market expectations for this product continue to rise. Analysts believe that the combination of Nvidia's extensive AI algorithm expertise and Foxconn's large-scale manufacturing capabilities has the potential to address the industry's pain points of humanoid robots: the difficulty of implementing the technology and the high cost of mass production. The introduction of "physical AI" may reshape robotics technology standards, driving the industry's evolution from "single-function devices" to "universal intelligent entities."
Industry experts predict that if this debut is successful, 2026 could be the first year for the commercialization of humanoid robots. The Nvidia-Foxconn partnership will also provide a new model for the cross-industry integration of technology and manufacturing.