NVIDIA is embarking on a significant domestic manufacturing initiative that will see the production of advanced AI supercomputers entirely on U.S. soil. This move represents a milestone in the company’s efforts to integrate cutting-edge technology with robust domestic infrastructure.
Collaborating with several industry leaders, NVIDIA has secured more than one million square feet of manufacturing space. These facilities, located in key regions across the country, will support the production of the company’s innovative Blackwell chips in Arizona and the assembly of state-of-the-art AI supercomputers in Texas.
Initial chip production is underway at a major facility in Phoenix, Arizona, managed by NVIDIA’s established manufacturing partners. Meanwhile, plans are in motion to construct dedicated supercomputer production plants in Texas, with Foxconn partnering in Houston and Wistron in Dallas. Both locations are expected to significantly increase their output over the coming months.
Given the complexities of the AI chip and supercomputer supply chain—which requires advanced manufacturing, packaging, assembly, and testing capabilities—NVIDIA is also working closely with key partners such as Amkor and SPIL to manage the packaging and testing operations in Arizona.
Over the next few years, the company aims to generate substantial investment in domestic AI infrastructure, potentially reaching a value of up to half a trillion dollars. These strategic partnerships not only enhance supply chain resilience but are also expected to create hundreds of thousands of jobs and drive long-term economic growth while expanding the global footprint of the involved companies.
CEO and founder Jensen Huang underscored the broader significance of this development, noting that producing AI supercomputers in the United States marks a transformative step in meeting the growing global demand for advanced AI technologies. NVIDIA plans to utilize sophisticated AI, robotics, and digital twin technologies—including platforms like NVIDIA Omniverse and NVIDIA Isaac GR00T—to design and efficiently operate these new facilities.
