Foxconn and Nvidia are reportedly in talks to introduce humanoid robots on the assembly floor of a new AI server production facility in Houston, Texas—a move that could mark a new era of automation in high-performance computing hardware.
If implemented, this would be the first time humanoids are involved in building Nvidia products, and also Foxconn’s first AI server line to integrate robot labor.
Sources told Reuters that deployment plans could finalize soon, with operations beginning as early as Q1 2026. Houston was selected strategically—offering greater space and a facility design optimized for automation.
While neither Nvidia nor Foxconn has officially confirmed the project, the move aligns with both companies’ ongoing push toward robotics. Foxconn has developed humanoid prototypes under its subsidiary, Foxconn Industrial Internet, training them for cable insertion, part placement, and light assembly.
At a recent event, Leo Guo, GM of Foxconn’s robotics division, revealed two robot models:
-
A bipedal robot with legs
-
A cost-effective wheeled version, mounted on an autonomous mobile platform
Pricing details remain confidential, but the wheeled model is expected to be cheaper and more viable for scaling.
Nvidia, meanwhile, powers much of the global humanoid robotics field with its AI platforms. At GTC 2025, CEO Jensen Huang predicted humanoid robots would be widespread in manufacturing within five years.
Nvidia’s broader Texas expansion includes a second site in Dallas, in partnership with Wistron, with production expected in 12–15 months.
But one key question remains: Where will these humanoid robots be built?
Foxconn could manufacture them in:
-
Taiwan (its robotics R&D base)
-
China (where it collaborates with UBTech)
-
Vietnam or India (for cost efficiency)
-
Or even Japan or South Korea, by partnering with advanced robotics firms
Whether the future of robotic assembly will rely on global supply chains or evolve into localized US-based production remains to be seen.
