(TAIPEI, TAIWAN) 29 October 2025 — In 2025, Wistron revealed its plans to invest $761 million in building AI supercomputing facilities in Texas. These state-of-the-art facilities are anticipated to become operational by the first half of 2026. As Wistron strengthens its AI-related product capabilities and advances its strategic roadmap, establishing manufacturing operations in the United States is a critical step in meeting the needs of our customers and advancing our global vision.

 

Wistron's Physical AI Strategy in U.S. Operations
Wistron will leverage NVIDIA technology to build its smart factories in the U.S., specifically for manufacturing AI supercomputers in Texas. The core of our strategy will revolve around physical AI, primarily through the use of digital twins enhance efficiency, design, and operations. 
 

Wistron’s implementation of physical AI, powered by NVIDIA AI infrastructure, will be focused on creating highly optimized, energy-efficient, and rapidly deployable manufacturing facilities:

 

• Digital Twins and NVIDIA Omniverse:
  • Wistron will leverage NVIDIA Omniverse libraries to build high-fidelity, physically accurate digital twins (virtual replicas) of our factories, including our new AI supercomputer assembly and test facilities in Texas.
  • These digital twins will be utilized for factory planning and operations, allowing engineers to simulate and optimize layouts, test equipment placement, and refine workflows before any physical construction begins, thus reducing time-to-market and avoiding costly change orders.
  • For our new NVIDIA supercomputer test rooms, the digital twin will help streamline facility planning and remote collaboration.
• Physics-Informed AI for Optimization:
  • Wistron will utilize frameworks like NVIDIA PhysicsNeMo to create physics-informed AI models. These models will accelerate complex simulations, such as Computational Fluid Dynamics (CFD), which are crucial for managing thermal dynamics.
  • By integrating these models into the digital twin, we will be able to accurately and rapidly predict airflow and temperature distributions in our supercomputing run-in test rooms. This will dramatically reduce simulation time and allow for near real-time control of the room's temperature to prevent overheating.
  • This thermal optimization is expected to result in significant energy efficiency improvements.
• Real-Time Operations:
  • The digital twin will be connected to thousands of physical sensors (IoT data) in the facility, providing a real-time digital dashboard of production status and operational metrics. This will enable managers to monitor performance, identify bottlenecks, and quickly allocate resources.
  • We will employ an AI-based recommendation system to identify the optimal testing locations for DUTs in our automated storage and retrieval systems (ASRS).

 

Onshoring AI Supercomputer Manufacturing in the U.S.
In collaboration with NVIDIA, this Texas plant will be one of the first in the U.S. to manufacture AI supercomputers entirely domestically, marking a critical milestone in the onshoring of technical supply chains for the growing AI industry.

 

Through its physical AI initiative and long-term investment in the United States, Wistron is reinforcing its leadership in AI innovation while fostering a future of smarter, more sustainable manufacturing processes.