NVIDIA Expands CUDA-X AI Tools For Semiconductor Manufacturing, Delivering 2–3x Faster Atom-Level Simulations And 30x Faster Chipmaking Simulations

NVIDIA has introduced the cuPotential library, part of its expanded CUDA-X AI tools, which enables developers to create AI agents for simulating atomic interactions, achieving 2-3x faster performance in early testing. Additionally, the NanoVDB library accelerates chipmaking simulations by 30x for AI-defined fabs, utilized by companies like Lam Research and KLA. This expansion aims to enhance semiconductor manufacturing efficiency amid rising demand for advanced chips, with various industry leaders adopting NVIDIA's technologies to improve production processes.