on may 28, nvidia ceo jensen huang, while attending the groundbreaking ceremony for the company’s new headquarters in taipei, clearly stated that taiwan is not only a critical fulcrum of the global ai hardware ecosystem but also the central engine driving the entire ai computing value chain. he announced that the company’s annual investment in taiwan will surge to us$150 billion, marking taiwan’s official emergence as a core hub for building next-generation artificial intelligence infrastructure.
according to reuters, citing his remarks at the event, nvidia’s annual investment in taiwan has skyrocketed from us$10–15 billion four to five years ago to the current level of us$100 billion, with plans to increase it by another 50% over the next few years, reaching us$150 billion annually. huang emphasized that key links across the entire supply chain—from advanced chip design and cutting-edge packaging technologies to the manufacturing of complete ai supercomputers—have been firmly established in taiwan, transforming the island from a mere contract manufacturing base into a strategic stronghold for ai computing, integrating r&d, mass production, and system integration under one roof.
the groundbreaking ceremony carried special significance—jensen huang attended alongside his parents, spouse, and children, underscoring both personal ties and unwavering strategic resolve. according to cnmo technology, construction of the new headquarters will begin within 2024 and is expected to be completed by 2030, creating approximately 4,000 high-value jobs and significantly strengthening nvidia’s collaborative technological capabilities with tsmc, while deepening vertical partnerships with leading ai server manufacturers such as foxconn, wistron, and quanta.
this round of intensive capital and technology investment is injecting strong momentum into taiwan’s semiconductor and electronics industries. trends such as tsmc’s release of advanced process capacity, foxconn’s accelerated shipments of ai servers, and wistron’s enhanced system integration capabilities are poised to accelerate further. notably, on may 21, amd also announced it would invest over us$10 billion in taiwan, focusing on building capabilities in ai chip packaging and final assembly, creating a resonance effect between these two industry giants.
south korean media analysis points out that, as nvidia and amd continue to ramp up their investments, taiwan is rapidly establishing a full-stack ai hardware hub spanning chip design, wafer fabrication, advanced packaging, system assembly, and software optimization. by contrast, although south korea holds a competitive edge in key storage technologies like hbm, it remains relatively weak in the broader ai hardware ecosystem and system-level coordination, making it difficult in the short term to challenge taiwan’s comprehensive leadership in ai computing infrastructure.
