NVIDIA introduced cuPotential, a new CUDA-X library for atomic-scale simulation. It accelerates GPU-optimized workflows by 2-3x in early testing. NanoVDB speeds chipmaking simulation up to 30x for Lam Research. NPP accelerates chip inspection for KLA. Applied Materials, Cadence, KLA, Lam Research, Siemens, Synopsys and others adopt these tools. The libraries help automate engineering work across chip design and manufacturing. Synopsys uses cuPotential, cuEST, and cuDSS for faster semiconductor simulations. Cadence integrates CUDA-X libraries into electromechanical system design. COMSOL achieves up to 10x faster multiphysics simulations with cuDSS. Voltair compresses complex simulation workflows from months to a day. cuPotential is available in early access. NanoVDB is open source within OpenVDB. NPP is available via the CUDA Toolkit. Lam Research targets NanoVDB-accelerated SEMulator3D release by year-end.
What changed
Semiconductor simulation and inspection ran on CPU-optimized or basic GPU workflows without specialized acceleration libraries.
What it unlocks
Building AI agents that simulate atomic interactions and accelerate chip design, manufacturing, and inspection workflows in AI-defined fabs.
- 2-3x faster atomic-scale simulation with cuPotential
- 30x faster wafer-etch modeling with NanoVDB
- 50x faster NPP NCC kernel vs CPU
- 85x faster probe-maintenance inference
What you need to act on it
- NVIDIA CUDA-X libraries
- GPU-accelerated workflows
Sources