NVIDIA has announced plans to manufacture its AI supercomputers entirely within the United States for the first time. The company is collaborating with manufacturing partners to establish production and testing facilities across Arizona and Texas.
Production of NVIDIA’s Blackwell AI chips has commenced at Taiwan Semiconductor Manufacturing Company (TSMC) facilities in Phoenix, Arizona. In parallel, NVIDIA is developing supercomputer manufacturing plants in Texas, working with Foxconn in Houston and Wistron in Dallas. According to the company, mass production at these sites is scheduled to scale over the next 12 to 15 months.
The company is also engaging with Amkor Technology and Siliconware Precision Industries Co., Ltd. (SPIL) for chip packaging and testing operations in Arizona. Collectively, these efforts involve over one million square feet of manufacturing space.
NVIDIA stated that over the next four years, it expects to manufacture up to $500 billion worth of AI infrastructure in the United States through its partnerships with TSMC, Foxconn, Wistron, Amkor, and SPIL. The move is positioned as part of a broader effort to expand domestic production capabilities and strengthen supply chain resilience.
The AI systems under production are described as forming the foundation of new data centers designed specifically for artificial intelligence applications. These data centers, referred to by NVIDIA as “AI factories,” are intended to support a growing number of high-power computing operations.
NVIDIA plans to integrate its proprietary technologies, including Omniverse for digital twin simulation and Isaac GR00T for robotics automation, to assist in the design and operation of the new manufacturing sites.
