Wistron's US AI server factory begins production, accelerating the creation of next-generation AI computing platforms for Nvidia (NVDA.US)

According to Zhitong Finance APP, Wistron Corporation, a technology service provider from Taiwan, announced on Wednesday that its first AI server manufacturing base in Fort Worth, Texas, USA has officially started operations. The facility is dedicated to producing next-generation AI computing platforms for Nvidia (NVDA.US). Following the announcement, Wistron's share price on the Taiwan Stock Exchange soared by 9.7%, highlighting the market's strong recognition of this strategic move.

Factory Overview: $700 million bet on U.S. domestic AI computing capacity, Annual Production of GB300 and Vera Rubin Superchips

The base, named the "D1 AI Smart Factory," covers 324,000 square feet and represents an investment of $700 million. The factory has already begun production of the Nvidia GB300 Grace Blackwell Ultra superchip platform—currently one of the most advanced AI processing platforms available. In the future, production will also expand to include Nvidia's Vera Rubin superchips, which are regarded by the industry as a critical engine driving the next generation of AI computing.

The completion of Wistron's D1 factory marks a pivotal upgrade for the U.S. domestic AI server supply chain. Previously, the main AI server production bases were located in Asia, and the U.S. lagged behind in system assembly and testing capabilities. Wistron's Fort Worth base will directly fill this gap.

Simon Lin, Wistron's Chairman, stated in a press release: “As we build AI infrastructure in the United States, this base will become one of the most important facilities in the coming years.” Wistron has indicated that the company is establishing a new model for AI infrastructure production based on digital manufacturing, energy optimization, and localized operations. The factory will expand U.S. domestic capabilities in system assembly and testing; these systems will be integrated into Nvidia’s DSX infrastructure to allow for large-scale deployment of highly efficient AI factories.

Nvidia CEO Jensen Huang commented: “As the engine of the next industrial revolution, demand for AI factories is astonishing, making it essential to manufacture in various locations around the globe.”

Wistron’s D1 plant is located in Fort Worth, Texas, an area that leverages the advantages of the U.S. domestic supply chain and logistics network. Fort Worth is an important transportation hub and industrial center in Texas and has attracted significant technology manufacturing investment in recent years. With demand for AI servers surging, limited U.S. assembly capacity had previously become a supply bottleneck. Wistron's commencement of production is expected to help relieve this pressure.

Strategic Significance: A Turning Point for the "Domestic U.S." AI Computing Supply Chain, Accelerating Server Production

The launch of Wistron's Fort Worth factory signals a structural shift in the global AI computing supply chain, transitioning from “offshore manufacturing” to “onshore manufacturing.”

Previously, the core components for AI servers—including GPU modules, motherboards, and cooling systems—were largely manufactured and pre-assembled in Asia before being shipped to the U.S. for final integration. While this model has advantages in logistics efficiency and cost, it faces shortcomings in supply chain resilience and delivery speed. Wistron’s U.S. factory will fundamentally change this paradigm—for the first time, the complete chain from chip to system to AI factory will be seamlessly connected within the U.S.

For Nvidia, Wistron's U.S. production capacity represents a strategic upgrade of its AI computing supply chain. The GB300 and Vera Rubin platforms are Nvidia’s core arsenal in meeting the surging demand for AI training and inference; their stable supply is directly crucial to the advancement pace of the global AI industry.

Wistron’s Fort Worth factory is not an isolated case. Following TSMC's (TSM.US) announcement last week of an additional $100 billion investment to expand capacity in Arizona, Wistron’s latest move highlights the broader trend of Taiwan's technology supply chain accelerating its transition to the United States. This is mainly because key clients like Nvidia have explicitly required AI computing capacity to be built in the U.S. to ensure supply chain resilience and delivery speed.

In addition, amid continued high tariffs maintained by the Trump administration, local U.S. manufacturing can effectively avoid tariff costs. The CHIPS and Science Act offers tax credits and subsidies, providing substantial financial support for establishing factories in the U.S.

Nvidia's manufacturing partner network now covers 43 states across the U.S., with key partners including Wistron, TSMC, and Foxconn. By 2026, these collaborations are expected to contribute around $485 billion to the U.S. GDP.