Cadence and NVIDIA Expand Collaboration to Redefine Engineering in the AI Era

[Erkan Teskancan]

Cadence and NVIDIA are expanding their partnership in AI-powered accelerated computing, combining physics-based simulation, digital twins, and agentic AI applications. This collaboration aims to accelerate engineering processes and increase efficiency in semiconductor design, physical AI systems, and large-scale AI factories.

### Partnership and Goals
Announced at CadenceLIVE Silicon Valley 2026, the partnership integrates Cadence's agentic AI-powered design, electronic design automation (EDA), and system design and analysis (SDA) technologies with NVIDIA's CUDA-X, AI physics, and Omniverse libraries. This will boost engineering productivity across three core design areas, accelerating innovation at true agent speed.

Cadence CEO Anirudh Devgan stated that this collaboration will accelerate the connection between design and physical reality, delivering unprecedented speed, accuracy, and confidence in simulation and system development. NVIDIA CEO Jensen Huang emphasized that CUDA-accelerated computing and AI are reshaping engineering processes, offering the opportunity to test and optimize everything as a full digital twin in the digital realm first.

### Accelerating Cadence Tools
The Cadence Millennium M2000 supercomputer, powered by NVIDIA CUDA-X, AI-physics, and Omniverse libraries, is accelerating Cadence's EDA and SDA solutions. This development provides up to a 100x speedup in engineering workflows. Leading customers such as Ascendence, Argonne National Laboratory, Honda R&D, Samsung, and SK Hynix are leveraging these accelerated solutions to bring their products to market faster.

### Next-Generation Chip Design with AgentStack
Cadence's ChipStack AI Super Agent has achieved up to a 10x increase in efficiency for semiconductor RTL design and verification using agentic AI and EDA tools. The newly introduced AgentStack extends this technology to physical design, custom/analog design, and system-level, coordinating multi-agent workflows with NVIDIA accelerated computing.

NVIDIA, as an early adopter of AgentStack, is supporting the system's development and scaling with real-world data. This approach reduces iteration times from days to hours with agent-based workflows that can reason across design hierarchies and protocols.

### Embedded Agentic AI for Physical AI
The partnership also includes the development of embedded agentic AI for physical AI. Cadence's Physical AI Stack, combined with NVIDIA's robotic simulation libraries and accelerated computing infrastructure, speeds up the "simulation-to-reality" transition for robots and autonomous machines.

This integrated workflow covers world model training, high-fidelity physics simulation, large-scale scenario testing, and real-time feedback. Monitoring and improvement with live digital twins are made possible in NVIDIA Jetson robotics and edge AI systems.

### AI Factory Digital Twins
The collaboration focuses on designing and optimizing digital twins of large-scale AI factories with NVIDIA Omniverse DSX Blueprint integration. These digital twins enable simulation-based system design to improve critical efficiency metrics, such as tokens processed per watt.

In a sample 10-megawatt AI factory use case, optimizing GPU power settings resulted in a 17% increase in energy efficiency per token and billions of dollars in additional revenue opportunities. Furthermore, increasing cooling system temperature yielded 32% more token efficiency. These high-fidelity digital twins allow factories to maximize performance within power and thermal limits.

### CadenceLIVE 2026 Demonstration
At CadenceLIVE 2026, Cadence will showcase its accelerated solutions, AgentStack, Physical AI Stack, and AI factory digital twins, demonstrating how it supports engineering teams from early concepts to deployment.

### About Cadence
Cadence, a leader in AI and digital twins, develops computational software to accelerate innovation in silicon and system engineering. It serves a broad market and industry portfolio, from next-generation chips to complete electromechanical systems. In 2024, it was named one of the world's top 100 best-managed companies by the Wall Street Journal.