Nvidia CEO Jensen Huang pledged to invest $150 billion (approximately 225 trillion won) annually in Taiwan, his hometown, during a visit to the island.
According to Reuters on Thursday, Huang made the announcement at a groundbreaking ceremony in Taipei for "Constellation," Nvidia's next-generation artificial intelligence (AI) development headquarters under construction. "Taiwan's annual investment will grow from $10 billion to $15 billion four to five years ago to as much as $150 billion going forward," Huang said.
Huang lavished praise on Taiwan, calling it "the epicenter of the AI revolution." He said, "This is where semiconductor chips, packaging and systems are made, and where AI supercomputers are born."
Huang's visit to his hometown — he was born in Tainan, Taiwan — is seen as a move to deepen ties with Taiwan as Nvidia faces limits in its China business. Constellation, Nvidia's Taiwan headquarters unveiled that day, is targeted for completion in 2030. Once finished, it is expected to employ 4,000 people, four times the size of Nvidia's current 1,000-strong workforce in Taiwan.
According to Taiwanese daily United Daily News, Huang signaled an active willingness to invest in Taiwanese companies, saying, "If the opportunity arises, I would be happy to invest." After meeting TSMC Chairman C.C. Wei and co-Chief Operating Officer Y.P. Chyn the previous day, Huang hosted a dinner that day for senior Quanta Computer executives including Chairman Barry Lam and Vice Chairman C.C. Yang. The TSMC dinner was attended by Debora Shoquist, Executive Vice President of Operations, who oversees the global supply chain, while the Quanta Computer dinner was attended by Raymond Teh, Vice President in charge of Asia-Pacific operations. Teh also serves as acting head of Nvidia's Taiwan branch.
The industry views Nvidia as likely to prioritize parts suppliers for AI data centers, which are facing supply shortages, or partners with whom it can build long-term cooperative relationships. The back-to-back meetings with TSMC, the world's largest contract chipmaker, and Quanta Computer, a global AI server original equipment manufacturer (OEM), suggest that Nvidia views OEM firms as key partners on par with TSMC and as strong investment candidates. Quanta Computer is also reported to have secured large-scale AI server orders from major North American communication service providers (CSPs) through Nvidia's AI platform.
"The cooperation schedule between the two companies in the second half of this year is expected to be very busy," Huang said, expressing gratitude to Quanta Computer. Huang, who said AI server production would reach an all-time high, also expressed confidence in the mass production of Nvidia's next-generation AI platform Vera Rubin in the second half of the year, United Daily News reported.
