NVIDIA announced its commitment to begin manufacturing AI infrastructure solutions in the United States. For the first time, the company will produce entire AI supercomputers — including its powerful new Blackwell chips — within the United States. Working with partners TSMC, Foxconn, Wistron, Amkor, and SPIL, NVIDIA plans to manufacture, package, assemble, and test AI systems across facilities in Arizona and Texas, anchoring a domestic supply chain capable of producing up to half a trillion dollars’ worth of AI infrastructure over the next four years.
The company confirmed that production has begun on NVIDIA Blackwell chips at TSMC’s fab in Phoenix, while Foxconn and Wistron are building out supercomputer manufacturing plants in Houston and Dallas, respectively. Mass production is expected to scale in 12–15 months. The company is also deploying its own AI, robotics, and digital twin platforms — including Omniverse and Isaac GR00T — to automate and optimize factory design and operations.
- Domestic AI Manufacturing at Scale: NVIDIA will produce AI chips and supercomputers entirely in the U.S., targeting high-volume deployments.
- Gigawatt-Scale AI Factories: These next-gen data centers are designed solely for AI processing, requiring enormous power and low-latency, high-throughput interconnects.
- Advanced U.S. Facilities: Over 1 million square feet commissioned for AI chip and system production, packaging, and testing in Arizona and Texas.
- Partners in Silicon and Systems: TSMC (chips), Foxconn and Wistron (systems), Amkor and SPIL (packaging/testing) all scaling U.S. operations with NVIDIA.
- Resilient AI Supply Chain: NVIDIA’s move improves national supply security and reshapes AI hardware geopolitics.
- Digital Twin & Robotics Integration: NVIDIA Omniverse and Isaac GR00T will drive factory automation and efficiency.
- Job Creation and Economic Impact: The initiative is expected to generate hundreds of thousands of jobs and drive trillions in long-term economic activity.
- Data Center Implications: These factories will require massive networking build-outs—from chip-to-chip interconnects to optical fabrics and energy-aware AI cluster designs.