Wistron Adopts NVIDIA Omniverse DSX to Accelerate Digital Factory Operations

Estimated reading time: 2 minutes

Wistron is actively implementing the NVIDIA Omniverse DSX Blueprint and Physical AI technology to drive the upgrade of AI factory smart manufacturing and a leap in operational efficiency. The company is focusing on optimizing high-energy-consuming testing areas and core facility systems within AI factories. These innovations are being applied across global manufacturing sites including Hsinchu, Taiwan, and the U.S. This initiative comprehensively connects the entire AI server manufacturing process—from initial planning and design to final mass production—ushering in a new era of smart manufacturing.

Utilizing Virtual Validation and Ecosystem Collaboration During the Construction Phase

During the plant construction stage, Wistron leverages OpenUSD technology to build high-fidelity virtual assets and employs digital twin technology to simulate 3D factory layouts and equipment in a virtual environment. The system features  integrates of thermal analysis with NVIDIA PhysicsNeMo AI physics models, significantly enhancing design validation efficiency. Furthermore, it connects with third-party professional software within the ecosystem for critical physical validation, ensuring that simulated environments are highly consistent with real-world sites.

Based on this foundation, Wistron has linked multiple global plants, drastically reducing the time required for complex multi-task layout analysis and validation, showcasing its robust global cross-regional simulation capabilities. This model also preemptively identifies and resolves production and cooling bottlenecks, shortening later-stage tuning times and accelerating the launch of new factories.

Implementing Smart Operating Systems and Zero-Hotspot Management During the Operational Phase

Once moving into the mass production and operational phase, Wistron established the AIF-DT (AI Factory Digital Twin) smart operating system. This system utilizes physics-aware inference technology to provide real-time analysis of power, thermal flow, and equipment status. It dynamically optimizes energy allocation and testing schedules to comprehensively improve factory operational efficiency.

Under this architecture, Wistron uses AIF-DT to intelligently recommend energy-saving solutions and thermal flow predictions, ensuring "zero-hotspot" operations and precise alignment with critical plant facility operations. This move has significantly reduced total factory energy consumption and power demand, bringing substantial energy-saving benefits to manufacturing sites worldwide.