Cadence and NVIDIA Redefine Engineering in the AI Era by Expanding Collaboration

[Erkan Teskancan]

Cadence and NVIDIA have expanded their collaboration to reshape engineering in the era of AI-powered accelerated computing. This partnership aims to achieve new levels of productivity, from semiconductors to physical AI systems and AI factories, by integrating agentic AI, physics-based simulation, and digital twin technologies.

### Foundations of the Collaboration
Cadence is combining its leadership in agentic AI-powered design, electronic design automation (EDA), and system design and analysis (SDA) with NVIDIA's CUDA-X, AI physics, and Omniverse digital twin libraries. This integration accelerates engineering productivity across three critical design areas, bringing innovations to market at real-time speeds.

Anirudh Devgan (Cadence CEO) emphasized that this partnership accelerates design and physical implementation processes, offering faster, more accurate, and reliable solutions in simulation and system development. NVIDIA CEO Jensen Huang stated that CUDA-accelerated computing and AI are fundamentally transforming engineering processes.

### Cadence Tools Accelerated by NVIDIA Technology
Cadence is accelerating its EDA and SDA solutions through NVIDIA CUDA-X, AI-Physics, Omniverse libraries, and the Millennium M2000 supercomputer. This enables speedups of up to 100x in engineering workflows. Customers like Honda, Samsung, and SK Hynix are already using these accelerated solutions to bring their products to market faster.

### AgentStack: Agentic AI for Next-Generation Chip Design
Cadence achieved a 10x increase in efficiency for RTL-level chip design and verification with its ChipStack AI Super Agent. The new AgentStack has been developed as a super agent that manages all stages of semiconductor and system design. AgentStack manages workflows from physical design, custom/analog design, and up to the system level, running on NVIDIA's accelerated computing infrastructure.

NVIDIA is testing the integration of this agent into real-world processes, which will accelerate the transition from script-based flows to agent-based flows.

### Embedded Agentic AI for Physical AI
Cadence and NVIDIA are integrating robotic simulation and high-fidelity multi-physics simulations for physical AI systems. This combination reduces the simulation-to-reality gap encountered by robots and autonomous machines in the real world.

This workflow provides a complete cycle from training to validation and real-world feedback. Simulations created with NVIDIA Isaac and Cosmos platforms are run on real-time digital twins and transferred to Jetson robotic systems.

### Cost and Performance Optimization with Digital Twins in AI Factories
The partnership is also effective in the design and simulation of AI factories. Cadence is developing digital twins for large-scale AI factories using NVIDIA Omniverse DSX Blueprints. These twins optimize the number of tokens processed per unit of power consumption, thereby increasing energy efficiency.

Specifically, GPU power settings, cooling, and system configurations are tested before simulations to make systems more efficient. For example, in a 10 MW facility usage scenario, a low-power mode (MaxQ) achieved 17% higher token efficiency and billions of dollars in potential annual additional revenue.

### Unveiling at CadenceLIVE 2026
Cadence will showcase accelerated solutions, AgentStack, the Physical AI Layer, and AI factory digital twins at CadenceLIVE 2026, thanks to this expanded collaboration with NVIDIA. It will demonstrate how engineers can manage their processes faster and with greater confidence, from early stages to implementation.

### About Cadence
Cadence is a market leader in AI and digital twins for silicon and system engineering. With its Intelligent System Design strategy, it enables semiconductor and system companies to design next-generation products. In 2024, it was recognized by the Wall Street Journal as one of the world's top 100 best-managed companies.