Taiwanese electronics maker, Wistron has revealed that its new AI hardware production facilities for its client and chip manufacturer Nvidia will be completed and ready for use by next year.
This comes as both firms are spearheading a landmark shift in US-based AI hardware production, promising to bring the manufacturing of servers and supercomputers home.
With new plants under construction in Texas and Arizona, both companies are aligning their timelines to meet the rising demand for AI infrastructure as well as reshaping global supply chains.
The company wants to align with its customer, Nvidia
Already, Nvidia has embarked on its complete US-based AI supercomputer manufacturing initiative, commissioning over a million square feet across Arizona and Texas. In Phoenix, its Blackwell AI chips have already entered production at Taiwan Semiconductor Manufacturing Company’s (TSMC) facilities.
Concurrently, Nvidia is collaborating with Foxconn in Houston and Wistron in Dallas to assemble and test AI servers and supercomputers. Both Texas sites are slated to begin mass production within 12 to 15 months, fully aligning with Nvidia’s plan to build up to $500 billion of AI infrastructure in the United States over the next four years.
CEO Jeff Lin confirmed that the Dallas facility, like Nvidia’s Texas plant, will be operational next year, in line with Nvidia’s timeline, and is already under discussion with additional potential clients, though names remain under wraps.
“I believe it’s exactly in line with what the customer announced. All our progress will follow the customer’s lead.”
Lin.
The initiative also marks a significant pivot from decades of offshoring cutting-edge chip and server assembly, bringing advanced packaging, assembly and testing closer to Nvidia’s primary market.
Wistron’s board also approved a $500 million investment in its new US subsidiary to support business development and strategic growth in the US.
Wistron and Nvidia target tariff-free production
Lin also emphasized that demand outside China remains robust despite US export controls on advanced chips, noting strong interest from markets in the Middle East and beyond.
Beyond meeting domestic demand, both Wistron and Nvidia are eyeing tariff-free production under the United States–Mexico–Canada Agreement. Wistron, in particular, is evaluating notebook assembly in Mexico to circumvent looming US tariffs on imports from several trading partners.
Meanwhile, Nvidia’s strategic partnerships extend to Amkor and SPIL in Arizona for chip packaging and testing operations, ensuring a seamless, high-precision supply chain.
The economic implications are staggering. As tens of “gigawatt AI factories” emerge in the coming years, analysts predict the US will see hundreds of thousands of new manufacturing jobs and trillions in economic activity fueled by home-grown AI hardware production.
Nvidia plans to leverage its own AI, robotics and digital twin solutions, from NVIDIA Omniverse for factory modelling to Isaac GR00T for robotic automation, to optimize operations and maintain cutting-edge efficiency.
And through diversifying its manufacturing footprint across leading global partners, TSMC for chip fabrication, Foxconn and Wistron for server assembly, and Amkor/SPIL for packaging, Nvidia is mitigating geopolitical risks and potential bottlenecks.
This US-centric approach not only aligns with federal incentives to onshore critical technologies but also sets a precedent for the broader semiconductor and electronics industry, which faces mounting pressures to localize production.
Looking ahead, Wistron’s engagement with new customers and its strategic choice to consider Mexican notebook production underscore a nimble response to shifting trade policies.
Meanwhile, the UAE’s recent agreement with Nvidia to build a massive AI campus outside the US, potentially consuming 500,000 of Nvidia’s top-tier AI chips annually from 2025, also exemplifies the global scale of this US manufacturing renaissance.
Cryptopolitan Academy: Coming Soon - A New Way to Earn Passive Income with DeFi in 2025. Learn More
