PartnershipJun 1, 2026, 01:02 AM
NVIDIA AI & Accelerated Computing Power TSMC Semiconductor Manufacturing
AI Summary
NVIDIA announced that TSMC is leveraging its accelerated computing and AI technologies to enhance semiconductor design and manufacturing processes. This collaboration aims to improve turnaround time, energy efficiency, yield, and operational productivity in advanced fabs. TSMC is utilizing NVIDIA CUDA-X libraries, AI models, Metropolis, TAO Toolkit, and exploring Omniverse for various applications including computational lithography, transistor simulation, defect inspection, and virtual fab environments.
Key Highlights
- TSMC adopts NVIDIA AI and accelerated computing for semiconductor design and manufacturing.
- NVIDIA CUDA-X libraries and AI models accelerate TSMC workloads across various processes.
- NVIDIA cuLitho improves computational lithography cost effectiveness/cycle time by 20-50%.
- NVIDIA cuEST enables 50x faster chemistry simulations for semiconductor material design.
- NVIDIA cuML accelerates large-scale analytics, significantly reducing process variation.
- NVIDIA Metropolis and TAO Toolkit enhance automated defect inspection at nanometer scale.
- TSMC explores NVIDIA Omniverse to build FabTwin for virtual fab environment and process optimization.
Price Impact
More from NVDA