Taipei, April 15 (CNA) Nvidia Corp. is teaming up with contract chipmaker Taiwan Semiconductor Manufacturing Co. (TSMC) and other partners to design and build facilities in the United States to produce the first-ever U.S.-made AI supercomputers.
In a statement posted on its blog, Nvidia said overnight that it is planned to produce up to US$500 billion of AI infrastructure products in the U.S. through its manufacturing partnerships over the next four years.
In addition to TSMC, Nvidia said, it is also working with Taiwan-based iPhone assembler and AI server maker Hon Hai Precision Industry Co. and Wistron Corp., another Taiwanese AI server supplier.
Also involved will be American IC packaging and testing services provider Amkor Technology Inc. and its Taiwanese counterpart Siliconware Precision Industries Co. (SPIL), Nvidia said.
Because many of the components that go into Nvidia’s AI graphic processing units are made in Taiwan, the shift in production to the U.S. could hurt Taiwan economically.
When asked about the issue at a legislative hearing, however, Economic Affairs Minister Kuo Jyh-huei (郭智輝) said on Tuesday that he could not comment on Nvidia’s AI supercomputer project because no details were available.
He did note, though, that the Ministry of Economic Affairs will closely monitor how Taiwanese companies invest overseas and provide necessary assistance.
Nvidia said the AI chip and supercomputer supply chain is complex and demands the most advanced manufacturing, packaging, assembly and test technologies and that its partners are deepening their ties with Nvidia to grow their businesses and expand their global footprint.
According to Nvidia, its advanced Blackwell AI chips have started production in TSMC’s wafer fab in Arizona, and it is building supercomputer production bases plants in Texas, with Hon Hai in Houston and with Wistron in Dallas.
Mass production at both plants is expected to ramp up in the next 12-15 months.
“The engines of the world’s AI infrastructure are being built in the United States for the first time,” Nvidia founder and CEO Jensen Huang (黃仁勳) said in the statement.
“Adding American manufacturing helps us better meet the incredible and growing demand for AI chips and supercomputers, strengthens our supply chain and boosts our resiliency.”
Nvidia’s announcement came at a time when U.S. President Donald Trump has been threatening tariff actions to force foreign manufacturers to invest in the U.S.
In the latest round of tariffs imposed by Trump on April 2, exports from Taiwan were to face a 32 percent import duty.
Trump announced a 90-day pause on the new measures last week, with a 10 percent duty to be applied instead to all countries except China.
TSMC pledged an additional US$100 billion in investment in the U.S. in March to build three more advanced wafer fabs, two IC assembly plants and one research and development center, which will push up its total investment in Arizona to US$165 billion.
