Nvidia CEO Jensen Huang unveiled a plan to pour $150 billion annually into a new headquarters in Taiwan, positioning the island as the core of the AI supply chain. He stressed that Taiwan supplies chips, packaging and the supercomputers that power today’s models, and that the investment will cement the region’s role through 2030. The facility, slated to break ground this year, will host R&D labs, testing centers and a logistics hub that ties directly into TSMC’s production lines.

The move follows Nvidia’s modest Taiwan spend of roughly $10‑15 billion a year in previous years and a 2024 push to produce AI chips in the United States under the Trump administration’s AI Action Plan. While U.S. output aims to meet half‑trillion‑dollar demand, advanced packaging still relies on Taiwanese fabs, prompting Huang to double down on local partnerships. Huang also flagged plans to collaborate with Foxconn, Wistron and Quanta, extending the Taiwan ecosystem beyond semiconductors into server assembly and AI infrastructure deployment.

Critics note the strategy runs counter to Trump’s goal of making America the AI hub, especially after export‑fee policies deterred China from buying Nvidia’s products. By anchoring operations in Taiwan, Huang hopes to safeguard supply lines for the upcoming Vera Rubin system and keep Nvidia’s market value climbing, a bet that could reshape the global chip ecosystem. Meanwhile, U.S. officials anticipate a pending review of additional data‑center chip tariffs, a move that could pressure Nvidia to balance its Taiwanese expansion with domestic compliance.