NVIDIA (NASDAQ:NVDA) announced that Taiwan Semiconductor Manufacturing Corporation (NYSE:TSM) is deepening their partnership by implementing the company’s suite of artificial intelligence and accelerated computing technologies across the entire semiconductor development and manufacturing process.
This effort spans multiple areas of chip manufacturing, from lithography and materials research to factory optimization and defect detection, as semiconductor manufacturers increasingly deploy AI-powered tools to improve efficiency and performance.
AI-powered lithography increases efficiency
One of the key technologies implemented is NVIDIA’s cuLitho platform, which TSMC uses for computational lithography applications.
According to the companies, this solution has improved either cost efficiency or processing cycle time by 20% to 50% compared to traditional CPU-based approaches.
The technology is designed to speed up one of the most computationally intensive steps in semiconductor manufacturing, helping optimize chip patterning and production workflows.
Accelerate materials research with faster simulations
TSMC also leverages NVIDIA’s cuEST software for electronic structure simulations, which can significantly speed up analysis of semiconductor materials.
The companies say the platform can perform chemical simulations up to 50 times faster than traditional methods and can support the design and development of advanced semiconductor materials.
By reducing simulation time, engineers can evaluate a wider range of material candidates and accelerate R&D cycles.
Machine learning enhances process control
To optimize manufacturing processes, TSMC has incorporated NVIDIA’s cuML machine learning library into its advanced process control system.
The platform enables analysis of hundreds of thousands of manufacturing parameters across thousands of production stages, enabling engineers to identify inefficiencies and more effectively reduce process variation.
According to TSMC, this technology has contributed to significant improvements in process consistency and operational performance.
GPU computing increases factory productivity
The semiconductor manufacturer is also deploying NVIDIA H200 GPUs to support production scheduling and factory management.
By using GPU-accelerated computing for scheduling calculations, TSMC can now better manage complex manufacturing constraints and optimize production flow within its manufacturing facilities.
The companies said these enhancements have resulted in measurable productivity gains across factory operations.
