NVIDIA Embeds CUDA-X Into Chip Fab Tools, Adding 2-3x Speedups
As seen on the 24/7 Wall St. homepage on October 14, 2026.
NVIDIA and Semiconductor Leaders Advance AI Across Chip Design and Manufacturing for AI-Defined Fabs
NVIDIA is now embedded in the tools its own suppliers use: Applied Materials, KLA, Lam Research, Cadence, Synopsys and Siemens are all running CUDA-X libraries across chip design, fab simulation and inspection. The new cuPotential library added 2-3x speedups on workflows already optimized for GPUs, extending the software moat into the fab equipment layer.
Continue ReadingShow less
Applied Materials, KLA, Lam Research, Cadence, Synopsys, and Siemens have all adopted NVIDIA's CUDA-X libraries, running them across chip design, fab simulation, and inspection workflows. These are the companies that make and verify the equipment NVIDIA itself depends on to produce its own chips.
The new cuPotential library is the sharpest edge of this announcement. It allows developers to build AI agents that simulate how atoms and molecules interact, and it delivered 2-3x speedups on workflows that were already GPU-optimized, meaning the gains come on top of a baseline that had already been tuned.
By embedding its software stack into the tools its own suppliers use, NVIDIA extends its software moat past the data center and into the physical manufacturing layer of the semiconductor industry. A competitor displacing NVIDIA at this layer would need to replicate the simulation and inspection software the entire fab ecosystem is now built around.
Investors tracking NVIDIA's software revenue story have focused on data center AI workloads, but this announcement signals that CUDA-X is becoming foundational to semiconductor manufacturing itself, a market with its own long upgrade cycle and switching costs that compound over time.
Mentioned: NVDA