Nvidia races to fulfill AI demand with its first Vietnam semiconductor hub

Nvidia is setting its sights on Vietnam with the company set to open a semiconductor design hub in the country in 2024.

During his first visit to Vietnam, Nvidia CEO Jensen Huang announced plans to deepen the company’s relationships with key Vietnamese partners including Viettel, FPT, Vingroup, and VNG, aiming to support Vietnam’s AI training and infrastructure, Reuters reported.

“Vietnam is already our partner as we have millions of clients here,” Huang was quoted as saying.

Recently, the White House announced a US-Vietnam Comprehensive Strategic Partnership focusing on technological and semiconductor advancements, including a novel Semiconductor Partnership and workforce development initiatives in Vietnam.

The country’s prime minister, Pham Minh Chinh, met with John Neuffer, CEO of the Semiconductor Industry Association, and US semiconductor company leaders in Ha Noi, emphasizing technology and innovation as key pillars in speeches about the countries’ relationship, a government media release said.

Vietnam’s investment minister Nguyen Chi Dzung has publicly expressed the country’s readiness with mechanisms and incentives to attract semiconductor and AI investment and requested Nvidia to consider establishing a research and development facility in Vietnam.

Vietnam has been a growing tech manufacturing destination for the past few years, and Nvidia management has said that the firm would be open to a third manufacturing partner in Vietnam.

“I think there are a lot of great foundry partners. TSMC has been a great one,” Nvidia Chief Financial Officer Colette Kress said during a recent UBS conference. “As you know, we also use Samsung today. Would we love a third one? Sure. We would love a third one.”

Vietnam is currently home to large Samsung and Intel facilities. Samsung said it plans to begin making chips in the country, while Intel – the likely third partner for Nvidia – has recently shelved plans to build a fab there because of reported concerns over power supply stability and excessive bureaucracy.

Nvidia’s expansion in Asia

Huang also recently paid a visit to Malaysia, where he announced a partnership with YTL Power to establish advanced AI infrastructure in Malaysia by mid-2024, featuring Nvidia H100 Tensor Core GPUs.

Huang is quoted as saying during the trip that he views Malaysia as a potential AI manufacturing hub.

Huang recently promised that more high-end GPUs for AI workloads are coming – and the company looks to be working with its partners to make that happen.

Local media in Taiwan report that Nvidia’s largest manufacturing partner, Taiwan Semiconductor, will open its new fab in Kumamoto, Japan in February. Production at the Kumamoto fab, a 12-inch fab with 16nm process technology, will begin in the second quarter of 2024.

TSMC is already looking ahead as it plans to have the next fab in the country manufacture chips on a 5nm process node, with a third fab making chips on a 3nm process node. 

TMSC founder Morris Chang met with Japan’s prime minister during the recent APEC conference in San Fransisco to discuss how the company could expand its presence in the country.  Currently, some customers are experiencing a lead time of up to 52 weeks for shipments of some of Nvidia’s most advanced GPUs.