As Foxconn gears up for a high-tech shift in US manufacturing, humanoid robots may soon assemble their and NVIDIA’s most powerful AI servers in Texas, says reports.
NVIDIA and Foxconn are planning to deploy humanoid robots to assemble cutting-edge AI servers in a US factory, according to a Reuters report. The project would mark the first time Foxconn uses humanoid robots in production, potentially transforming its Houston plant in Texas into a hub for AI server manufacturing.
The report indicates that the robots will assemble NVIDIA’s latest GB300 Blackwell servers, which are among the world’s most advanced AI platforms. The GB300 NVL72 system combines 72 Blackwell GPUs and 36 NVIDIA CPUs, making it a key component in the development of next-generation AI.
This shift is linked to broader trends in automation and US-based manufacturing. The Trump administration’s push for domestic production, combined with tariffs on Mexico, has prompted firms like NVIDIA and Foxconn to invest heavily in American facilities. Foxconn has reportedly committed $450 million to expand its operations in Houston.
Foxconn’s humanoid robots are being developed through a partnership with Chinese robotics firm UBTECH. However, it remains unclear whether these specific machines will be deployed in the US plant.
Meanwhile Tesla, which is also exploring humanoid robots for its factories, is expected to begin similar deployments by the end of 2025.
NVIDIA is increasingly involved in robotics, providing hardware and software platforms such as DGX, AGX, and Omniverse that are used in developing and training autonomous systems.
The company recently partnered with measurement technology firm HEXAGON to contribute to the development of AEON, a humanoid robot designed for industrial tasks.
The robots in Houston are expected to begin operations once development is finalised in the coming months. As demand for humanoid robots grows, NVIDIA stands to benefit through increased sales of its powerful AI chips, which are vital for robot control, learning, and simulation.
