The Zhitong Finance App learned that Wistron Corporation (Wistron Corporation), a Taiwanese technology service provider, announced on Wednesday that its first AI server production site in Fort Worth (Fort Worth), Texas, USA, has officially been put into operation to manufacture a next-generation AI computing power platform for NVDA.US (NVDA.US). After the news was announced, Wistron's stock price on the Taiwan Stock Exchange surged 9.7%, highlighting the market's high recognition of this strategic layout.
Factory panorama: $700 million bet on AI computing power production capacity in the US, annual production of GB300 and Vera Rubin superchips
The site, named the “D1 AI Smart Factory,” covers an area of 324,000 square feet and has an investment of $700 million. Currently, the plant has begun production of Nvidia's GB300 Grace Blackwell Ultra superchip platform — one of the most advanced AI processing platforms currently available. In the future, the factory will also expand production of Nvidia's Vera Rubin superchip, which is regarded by the industry as the key engine driving the next generation of AI computing.
The completion of Wistron's D1 factory marks a key upgrade in the US AI server supply chain. Prior to that, the main production base for AI servers was located in Asia, and system assembly and testing capabilities in the US were relatively limited. Wistron's Fort Worth base will directly fill this gap.
Simon Lin, chairman of Wistron, said in a press release: “As we build AI infrastructure in the US, this base will become one of the most important facilities in the next few years.” Witron said the company is establishing a new AI infrastructure production model based on digital manufacturing, energy optimization and localized operation. The plant will expand its ability to assemble and test systems in the US, and these systems will be integrated into Nvidia's DSX infrastructure to enable large-scale deployment of energy-efficient AI plants.
Nvidia CEO Hwang In-hoon said, “As the engine of the next industrial revolution, the market demand for AI factories is amazing, so production must be carried out everywhere.”
Wistron's D1 plant is located in Fort Worth, Texas. This location allows it to take full advantage of the US supply chain and logistics advantages. Fort Worth is an important transportation hub and industrial center in Texas, and has been attracting large investment in technology manufacturing in recent years. As demand for AI servers exploded, insufficient assembly capacity in the US once became a bottleneck limiting supply. Wistron's commissioning will help ease this pressure.
Strategic significance: the “American localization” shift in the AI computing power supply chain to accelerate server production
The commissioning of Witron's Fort Worth plant marks that the global AI computing power supply chain is undergoing a structural transformation from “offshore manufacturing” to “onshore manufacturing.”
Previously, the core components of AI servers — including GPU modules, motherboards, cooling systems, etc. — were mainly manufactured and initially assembled in Asia, and then shipped to the US for final integration. This model has advantages in terms of logistics efficiency and cost, but it has shortcomings in supply chain resilience and delivery speed. Wistron's US factory will fundamentally change this pattern — the complete chain from chips to systems to AI factories will be seamlessly connected in the US for the first time.
For Nvidia, Wistron's US production capacity means a strategic upgrade to its AI computing power supply chain. The GB300 and Vera Rubin platforms are Nvidia's core weapons to address the growing demand for AI training and inference, and their stable supply is directly related to the pace of evolution of the global AI industry.
Wistron's Fort Worth plant was not an isolated incident. Following TSM.US (TSM.US)'s announcement last week of an additional investment of 100 billion US dollars to expand production capacity in Arizona, Wistron's latest layout highlights the broad trend of Taiwan's technology industry chain accelerating its shift to the US. Mainly, core customers such as Nvidia clearly require the establishment of AI computing power production capacity in the US to ensure supply chain resilience and delivery speed.
Furthermore, as the Trump administration continues to maintain high import tariffs, domestic production in the US can effectively avoid tariff costs. The tax credits and subsidy policies provided by the CHIPS and Science Act (CHIPS and Science Act) also provided considerable financial support for setting up factories in the US.
Nvidia's manufacturing partner network spans 43 states, with core partners including Witron, TSMC, and Foxconn (Foxconn). These collaborations are expected to contribute approximately $485 billion to US GDP by 2026.