NVIDIA and Cadence have announced a major partnership expansion that could fundamentally reshape how engineers design chips and build AI systems, combining artificial intelligence agents with physics-based simulation to compress what once took days into hours. The collaboration integrates Cadence's design automation tools with NVIDIA's CUDA-X computing libraries and artificial intelligence physics models, creating an end-to-end workflow that lets engineers explore, test, and optimize designs at unprecedented speed .

What Are Agentic AI Agents and Why Do Engineers Need Them?

At the heart of this partnership is something called "agentic AI," which differs from the chatbots most people interact with daily. Rather than waiting for human commands, agentic AI agents can reason through complex problems, make decisions, and execute tasks autonomously across multiple steps. In semiconductor design, these agents can analyze chip architectures, identify optimization opportunities, and iterate through design variations without constant human intervention .

Cadence recently introduced its ChipStack AI Super Agent, which applies agentic AI combined with traditional electronic design automation (EDA) tools to transform how engineers design and verify semiconductor chips. Early deployments at more than 10 leading customers have already demonstrated up to a 10X productivity boost in their design and verification tasks . Building on this success, Cadence unveiled AgentStack, a head agent designed to orchestrate all aspects of semiconductor and system design, extending beyond just chip verification into physical design, custom analog design, and system-level workflows .

"Agentic AI and digital twins are reshaping the entire engineering landscape, from semiconductor design to planetary-scale AI systems," stated Anirudh Devgan, president and chief executive officer of Cadence. "Our expanded collaboration with NVIDIA accelerates the convergence of design and physical realization, connecting the Cadence AgentStack, Physical AI Stack, and AI factory digital twins with NVIDIA's breakthroughs in accelerated computing to deliver unprecedented speed, accuracy and trust in simulation and system development."

Anirudh Devgan, President and Chief Executive Officer, Cadence

How Can Engineers Use These New Tools to Speed Up Design?

• Accelerated Simulation Workflows: Cadence and NVIDIA are accelerating a wide range of principled solvers and leveraging AI physics models to deliver engineering workflows up to 100X speedup, allowing engineers to test thousands of design variations in the time it previously took to test dozens .
• Agent-Driven Design Flows: The partnership marks a significant shift from traditional script and graphical user interface-driven workflows to agent-driven flows capable of reasoning over design hierarchies, relationships, and protocols, dramatically compressing iteration cycles from days to hours .
• Physics-Based Digital Twins: By combining Cadence's high-fidelity multiphysics simulation with NVIDIA's robotics simulation libraries and accelerated computing, engineers can create accurate digital twins that help close the critical "sim-to-real" gap for robots and autonomous machines before deploying them in the physical world .

How Is This Changing AI Factory Design and Efficiency?

The partnership extends beyond chip design into a new frontier: optimizing massive AI factories where companies train and run large language models. These facilities consume enormous amounts of power, making efficiency a critical economic factor. Cadence and NVIDIA are creating digital twins of AI factories that help customers design, simulate, and optimize large-scale systems for training and inference workloads .

The focus is on a new metric called "tokens per watt," which measures how many model tokens a system can process for each unit of power consumed. In a joint 10-megawatt AI factory use case, modeling GPU operation at reduced power settings demonstrated up to 17% more tokens per watt and billions of dollars of incremental annual revenue per gigawatt for large-scale deployments . When combining reduced power operation with warmer coolant temperatures, the digital twins showed roughly 32% more tokens per watt, suggesting significant untapped efficiency gains .

"For the first time, we can innovate in the digital world, exploring, testing, and optimizing ideas at unprecedented speed and scale, by building everything as full-fidelity digital twins first," explained Jensen Huang, founder and CEO of NVIDIA. "Together, NVIDIA and Cadence are bringing this vision to life, transforming how engineers design, build and operate the world."

Jensen Huang, Founder and Chief Executive Officer, NVIDIA

Which Companies Are Already Using These Tools?

The partnership isn't theoretical. Cadence EDA and system design analysis customers and partners including Ascendence, Argonne National Laboratory, Honda R&D, Samsung, and SK Hynix are already leveraging Cadence solutions accelerated by NVIDIA to bring accelerated products to market faster . NVIDIA itself is adopting the AgentStack flow in its own semiconductor and system design workflows, providing real-world feedback that helps Cadence refine and scale the technology for broader industry deployment .

For physical AI systems like robots and autonomous machines, the collaboration integrates Cadence's Physical AI Stack with NVIDIA's Isaac open-source simulation libraries and Cosmos open-world models. This creates an end-to-end, agent-orchestrated workflow that links world-model training, accurate physics simulation, large-scale scenario testing, and continuous real-world feedback . The workflow spans virtual training in NVIDIA Isaac Sim and Isaac Lab, detailed physics evaluation through Cadence models, and mission-scale scenario simulation in VTD and VTDx environments .

The results of this expanded partnership represent a fundamental shift in how engineering happens. Rather than building physical prototypes and testing them in the real world, engineers can now build complete digital twins, run millions of simulations, and optimize designs using AI agents that reason through complex tradeoffs. This approach promises to accelerate innovation cycles across semiconductors, robotics, and AI infrastructure, while simultaneously reducing costs and improving safety before systems ever reach production .