NVIDIA and Wistron invest $700M in next-generation AI manufacturing
Subscribe to our free newsletter today to keep up to date with the latest manufacturing news.
The race to build artificial intelligence is often framed around software breakthroughs and increasingly powerful chips. Less visible, but just as important, is the manufacturing capacity required to turn those technologies into commercial systems at scale.
As demand for AI computing grows, the companies assembling the hardware behind advanced models are taking on a more strategic role in the technology economy. Their factories must handle complex components, demanding quality controls and production schedules shaped by a fast-moving market.
Wistron’s new manufacturing facility in Fort Worth, Texas, reflects that shift. Developed in partnership with NVIDIA, the site is designed to produce advanced AI computing platforms and show how simulation, automation and digital twins can change the way factories are planned and operated.
A new generation of manufacturing is forming around AI infrastructure
Wistron’s first US manufacturing facility represents a combined investment of $700 million and covers 324,000 square feet in Fort Worth. The factory has begun producing NVIDIA GB300 Grace Blackwell Ultra Superchips and is preparing to support production of Vera Rubin Superchips.
Output is expected to reach tens of thousands of AI computing boards each month in 2026. The site supports NVIDIA’s stated plan to produce up to $500 billion in AI infrastructure in the United States over four years.
The investment also has clear regional implications. More than 500 employees have joined the operation, and staffing is expected to reach about 1000 workers as production increases. A facility of this size can create demand for local suppliers, logistics providers, maintenance teams and engineering services.
AI infrastructure manufacturing differs from conventional electronics production in several ways. The systems combine advanced processors, high-speed interconnects, specialized cooling equipment and dense power systems. Each element must be assembled and tested to exact standards.
That complexity raises the cost of errors. A poorly designed production line can slow output, create quality problems and delay delivery schedules. Manufacturers must identify potential bottlenecks before physical equipment is installed.
This pressure is changing how factories are designed. Companies are moving away from disconnected production stages and toward systems in which engineering, operations, maintenance and quality teams share data across the entire manufacturing process.
For Wistron, that meant building the Fort Worth factory in a virtual environment before completing it in the physical world.
Digital twins are becoming central to factory planning and operations
One of the most notable features of Wistron’s Texas project is that much of the facility existed digitally before construction was finished.
Using NVIDIA Omniverse and other simulation tools, Wistron created a digital twin of the manufacturing operation. Engineers could test production layouts, material flows, robotics placement and operating scenarios before installing equipment on the factory floor.
This approach reduces uncertainty during development. Teams can identify inefficient layouts, equipment conflicts and workflow problems through simulation instead of discovering them after construction.
That can lower redesign costs and help a factory begin production faster. It can also give engineers more confidence when making changes to systems that would be expensive or disruptive to test in a live facility.
The digital twin remains useful after production begins. Data from the physical factory can be fed back into the virtual model, allowing teams to study performance, test scheduling changes and evaluate process adjustments without interrupting operations.
According to NVIDIA case-study data, Wistron reduced the time required for certain manufacturing simulations from about 15 hours to 3.6 seconds through accelerated computing. The company also reported energy savings of about 10% from AI-assisted optimization.
Synthetic data has improved inspection accuracy by 5% to 20% in selected manufacturing processes, according to the same case study. In these applications, computer-generated training data helps vision systems identify defects that may be rare or difficult to capture in real production environments.
The figures show why digital twins are moving beyond their earlier role as engineering models. They are becoming working systems that support decisions throughout a factory’s life cycle.
For manufacturers producing high-value equipment, the business case is direct. Faster simulations can shorten planning cycles. Better inspection can reduce waste. Energy optimization can lower operating costs. Virtual testing can limit disruption when production lines need to change.
These tools may become standard for companies building advanced electronics and AI systems, especially as product cycles become shorter and factories face pressure to scale more quickly.
The Fort Worth project points to a new model for AI supply chains
The Wistron facility also reflects a change in manufacturing strategy.
For decades, electronics companies concentrated production in global supply networks designed primarily around cost and efficiency. Recent disruptions have pushed companies to place more weight on resilience, regional capacity and access to critical components.
Expanding production in Texas gives NVIDIA and Wistron a stronger manufacturing presence in the United States. It also adds assembly and systems production to a domestic technology base that includes semiconductor fabrication, data center construction and energy infrastructure.
The emerging model is not based on moving every stage of production into one country or region. It is based on creating manufacturing clusters with enough capacity to reduce delays, respond to demand and limit exposure to disruptions.
Texas offers several advantages for that approach. The state has an established industrial base, major transportation links and a large pool of engineering and technical workers. It is also attracting continued investment in data centers, energy systems and computing infrastructure.
Those factors could make Fort Worth an important part of the supply network supporting AI deployment across the United States.
The project also highlights two parallel effects of AI on industry. First, AI is creating demand for a new class of computing equipment. Second, it is giving manufacturers new ways to design, test and manage the factories that produce that equipment.
Wistron’s facility shows how closely those two developments are connected. The factory is producing AI systems, but its own planning and operations also rely on AI-enabled simulation and automation. That relationship may become one of the defining features of advanced manufacturing. Companies will compete not only on the products they make, but also on how quickly they can design facilities, adapt production lines and improve output.
Source:
NVIDIA