NVIDIA and Semiconductor Leaders Advance AI Across Chip Design and Manufacturing for AI-Defined Fabs
nvidia ·
NVIDIA has announced that semiconductor leaders are adopting its CUDA-X libraries to advance the development of AI-defined fabs, which use artificial intelligence to optimize chip manufacturing processes. The company introduced cuPotential, a new library that accelerates atomic-scale simulations by helping AI agents model molecular interactions, achieving speedups of 2-3x in early testing. Other key tools include NanoVDB for 3D data processing and NPP for image analysis, which help partners like Lam Research and KLA significantly faster simulation and inspection workflows. Major industry players such as Synopsys, Cadence, Applied Materials, and Siemens are integrating these technologies to automate engineering tasks, improve yield, and reduce design times. For instance, Synopsys reports up to a 30x speedup in certain simulations, while Lam Research expects wafer-etch modeling to accelerate by 30x using NanoVDB. These advancements allow manufacturers to predict process behavior and manage complexity more effectively as demand for advanced chips rises. NVIDIA aims to support the industry's transition toward fully autonomous semiconductor engineering through these accelerated computing solutions.
nvidia