Cadence has expanded its collaboration with NVIDIA to develop integrated solutions combining agentic AI, physics-based simulation and digital twin technologies, aimed at advancing engineering workflows across semiconductor design, physical AI systems and large-scale AI infrastructure.
Cadence & NVIDIA Deepens Partnership
Per its press release, Cadence says the collaboration will integrate its electronic design automation (EDA) and system design and analysis (SDA) tools with NVIDIA’s CUDA-X platform, AI physics models and Omniverse libraries, enabling a unified approach to design, simulation and optimization across chip development, robotics and data center systems.
Cadence’s tools are being accelerated using NVIDIA computing platforms to improve simulation efficiency through its Millennium M2000 supercomputer running on aforementioned NVIDIA infrastructure (with a claimed 100x speedup). The partnership also includes the introduction of AgentStack, a system designed to coordinate AI agents across semiconductor and system design workflows, building on earlier ChipStack AI Super Agent deployments, with NVIDIA adopting the framework in its internal processes.
The collaboration extends to physical AI development by integrating Cadence tools with NVIDIA robotics simulation platforms, supporting workflows that link virtual training, simulation and real-world deployment. It also includes digital twin solutions for AI factory environments, enabling modelling of system configurations, power usage and cooling performance before deployment, with a focus on improving processing efficiency relative to energy consumption.
Pokdepinion: Big ambitions with AI from both parties.