Cadence Design Systems reported up to 10x productivity gains from AI-driven chip design tools. The results highlight accelerating integration of AI across semiconductor design, robotics, and infrastructure engineering.
The company announced expanded collaborations with Nvidia and new integrations with Google Cloud at its CadenceLIVE event. The Nvidia partnership combines Cadence’s physics-based simulation with CUDA-X libraries and Omniverse environments to model real-world system behavior. Separately, Cadence introduced an AI agent for chip layout tasks, delivered through Google Cloud using Gemini models.
Can Simulation Replace Real-World Engineering Trials?
The joint platform targets system-level design across semiconductors, robotics, and data center infrastructure. Engineers can simulate thermal, mechanical, and networking interactions before deployment, reducing reliance on physical prototyping. Nvidia said companies including ABB, FANUC, YASKAWA, and KUKA are already using simulation frameworks to test robotic systems virtually.
This reflects a broader shift toward digital twin environments and AI-assisted engineering. Training robotic systems in simulation reduces the need for costly real-world data collection, with outcomes tied to the accuracy of physics models. Compared with traditional workflows, simulation-driven design can compress development cycles and reduce infrastructure risks.
“The more accurate (generated training data) is, the better the model will be,” said Cadence CEO Anirudh Devgan.
Nvidia CEO Jensen Huang added that the collaboration extends to robotic systems, emphasizing integration between simulation and AI-driven control.
Cadence’s ChipStack AI platform coordinates tasks across multiple design stages, from circuit definition to physical layout. The company said cloud deployment allows teams to run complex workloads without relying on on-premise infrastructure. Nvidia also introduced its open-source Ising models, designed to improve quantum error correction with up to 2.5 times faster performance.
The next catalyst will be whether simulation-first engineering can scale across industries where system complexity and deployment costs continue to rise.
