NVIDIA and design software maker Cadence announced a major partnership expansion that combines agentic AI, physics-based simulation, and digital twins to dramatically accelerate engineering workflows across semiconductor design, robotics, and AI data centers. The collaboration leverages NVIDIA's CUDA-X accelerated computing platform and Cadence's electronic design automation (EDA) tools, which help engineers design computer chips and complex systems. Early deployments show engineering workflows running up to 100 times faster, with some AI factory simulations achieving 32% better energy efficiency .

What Are Agentic AI Agents and Why Do They Matter for Engineering?

Agentic AI refers to artificial intelligence systems that can autonomously reason through complex problems, make decisions, and execute tasks with minimal human intervention. Unlike traditional AI chatbots that respond to individual queries, agentic AI agents can orchestrate multiple steps across long workflows, learning and adapting as they go. In engineering, this means AI agents can handle entire chip design cycles, from initial concept through verification and physical layout, compressing what once took days into hours .

Cadence recently introduced its ChipStack AI Super Agent, which applies agentic AI combined with traditional EDA tools to transform semiconductor design and verification. Early deployments at more than 10 leading customers have already demonstrated up to a 10X productivity boost in design and verification tasks. Building on this foundation, Cadence unveiled AgentStack, a head agent designed to orchestrate all aspects of semiconductor and system design, extending beyond chip design into 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. 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," said Anirudh Devgan, president and chief executive officer of Cadence.

Anirudh Devgan, President and Chief Executive Officer, Cadence

How Are NVIDIA and Cadence Accelerating Three Critical Engineering Domains?

• Semiconductor Design: AgentStack connects Cadence agents with EDA platforms that leverage NVIDIA Nemotron language models and run on NVIDIA accelerated computing infrastructure. This enables agent-driven workflows that can reason over design hierarchies, relationships, and protocols, compressing iteration cycles from days to hours rather than requiring traditional script and graphical user interface-driven approaches .
• Physical AI and Robotics: Cadence and NVIDIA are combining the Cadence Physical AI Stack with NVIDIA robotics simulation libraries to close the critical "sim-to-real" gap for robots and autonomous machines. By integrating high-fidelity multiphysics simulation with NVIDIA Isaac simulation libraries and Cosmos open-world models, customers gain end-to-end agent-orchestrated workflows that link world-model training, accurate physics, large-scale scenario testing, and continuous real-world feedback .
• AI Factory Optimization: The collaboration extends to AI data centers, where Cadence integrates the NVIDIA Omniverse DSX Blueprint to enable next-generation AI factory digital twins. These digital twins help customers design, simulate, and optimize large-scale AI factories for training and inference, focusing on a critical new metric called tokens per watt, which measures how many model tokens are processed per unit of power consumed .

What Real-World Performance Gains Are Companies Seeing?

The partnership is already delivering measurable results. Cadence and NVIDIA are accelerating Cadence EDA and SDA (system design and analysis) solutions with NVIDIA CUDA-X, AI physics, and Omniverse libraries, with the Cadence Millennium M2000 Supercomputer powered by NVIDIA AI infrastructure. As part of this expanded collaboration, Cadence will accelerate its wide range of principled solvers and leverage AI physics models to deliver engineering workflows up to 100X speedup .

In a joint 10-megawatt AI factory use case, modeling GPU operation at reduced power settings (a feature called MaxQ) demonstrated up to 17% more tokens per watt and billions of dollars of incremental annual revenue per gigawatt for large-scale deployments. Digital twins of NVIDIA DSX-based AI factories have also demonstrated that combining MaxQ operation with warmer coolant could yield roughly 32% more tokens per watt. By capturing the interactions between IT load, cooling systems, airflow, and control logic in a high-fidelity digital twin, operators can safely push their AI factories toward maximum efficiency while respecting power and thermal constraints .

"CUDA-accelerated computing and AI are reinventing the engineering process. 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. Together, NVIDIA and Cadence are bringing this vision to life, transforming how engineers design, build and operate the world," said Jensen Huang, founder and CEO of NVIDIA.

Jensen Huang, Founder and Chief Executive Officer, NVIDIA

Which Companies Are Already Using These Tools?

Cadence EDA and SDA customers and partners are already leveraging Cadence solutions accelerated by NVIDIA to bring accelerated products to market faster. Early adopters include Ascendence, Argonne National Laboratory, Honda R&D, Samsung, and SK Hynix. NVIDIA itself is adopting the AgentStack flow in its own semiconductor and system design workflows, providing real-world feedback that will help Cadence harden and scale AgentStack for broader industry deployment .

This evolution marks a significant shift from traditional script and GUI-driven engineering flows to agent-driven flows that are capable of reasoning over design hierarchies, relationships, and protocols. The result is dramatically compressed iteration cycles, allowing engineers to explore more design variations and optimize systems faster than ever before. As AI factories become increasingly critical infrastructure for training large language models and other AI systems, the ability to simulate and optimize these facilities before physical deployment could save companies billions in operational costs .