- Foxconn prepares robots to perform manual tasks such as cable insertion and parts assembly
- Nvidia can become the first to use humanoid robots in its AI server production chain
- Foxconn’s Houston factory has been selected for its largest space and new adaptable layout design
Foxconn and Nvidia would have discussed the deployment plans of humanoid robots in a new installation of production of AI servers in Houston.
If it is implemented, it would be the first time that humanoids have been used in the assembly of Nvidia products and the first range of AI servers of Foxconn to incorporate them.
Familiar sources with the case, which refused to be appointed, told PK Press Club that the deployment could be finalized soon, the operations may start in the first quarter of 2026.
The new factories provision makes robotics more feasible
Houston was considered a strategic choice because of its new design of installations and its largest space available compared to existing AI servers manufacturing sites.
However, neither Nvidia nor Foxconn officially confirmed the plan, but the reports correspond to the growing interest of the two companies for automation and robotics, in particular in the fields which require speed and flexibility, such as the production of IA servers.
Foxconn has developed its own humanoid robots via its subsidiary Foxconn Industrial Internet and also causes robots to perform basic tasks such as cable insertion, object placement and light assembly work.
During a recent event in Taipei, Leo Guo, director general of the company’s robotic division, revealed that two types of robots will be presented in November 2025.
A version has legs, while the other is mounted on an autonomous wheeled mobile base, which, according to Guo, “would cost less than the version with the legs”, although it has refused to share specific figures.
Nvidia, for its part, has a solid participation in the field of humanoid robotics, because it provides AI platforms used by many developers to create humanoid machines.
During his recent NVIDIA GTC 2025 event, CEO Jensen Huang said the general use of these robots in manufacturing was “less than five years old”.
Nvidia’s wider plans in Texas include a second partnership in Dallas with Wistron, and the two sites should start production within 12 to 15 months.
However, an unanswered question persists: if Nvidia is pressure for humanoid robots in the United States, where will these robots be built?
Foxconn can turn to Taiwan, where its robotics unit is based, or to China, where it operates large -scale factories and has already collaborated with Ubtech.
Vietnam and India are also plausible choices, given the continuous expansion of Foxconn in these countries to reduce production costs.
Alternatively, the company could outsource components to robotics manufacturers established in Japan or South Korea, where technology is more advanced and well established.
Although the United States can be taken into account for localized production, this is probably part of a longer-term strategy.
This key detail will determine how evolving the approach is and if the future of automated manufacturing will be based on global supply chains or more localized development.
