Nvidia CEO Jensen Huang has announced that the company plans to invest hundreds of billions of dollars in U.S.-made chips over the next four years, signaling a shift away from reliance on Asian manufacturing amid rising geopolitical tensions and potential tariffs. Huang made the announcement during the 2025 GTC conference, confirming that Nvidia is already producing its Blackwell systems in the U.S. and working with TSMC’s Arizona facility for chip manufacturing.
In an interview with Reuters, Huang stated, “We are in it,” referring to the production of actual chips—rather than just prototypes—at TSMC’s Fab 21 in Arizona. While the exact volumes of production remain unspecified, Nvidia is committed to increasing its manufacturing footprint in the U.S., with Huang estimating a total investment of up to $500 billion for chips over the next four years, including hundreds of billions allocated to U.S. production.
The shift to U.S.-based production comes at a time when Nvidia’s data center GPUs are seeing significant growth, a trend that is expected to continue in the coming years. As the demand for high-performance computing and AI accelerates, Nvidia plans to expand its portfolio beyond GPUs, including CPUs, DPUs, networking chips, and SoCs for vehicles, all of which are being increasingly produced in American facilities.
The company also stands to benefit from the expansion of semiconductor production capacity in the U.S. by key suppliers like Micron, SK hynix, and Intel, who are all building new fabs to meet growing demand. With Nvidia increasingly sourcing components from American facilities, the company’s plans reflect a broader trend of reshoring semiconductor manufacturing.
While the total investment figures remain fluid, Nvidia’s commitment to bolstering U.S. production represents a significant step toward diversifying its supply chain and reducing dependence on international manufacturing hubs. The expansion of domestic semiconductor manufacturing could reshape the U.S. tech industry, particularly in sectors like data centers and AI infrastructure.
