Cadence and NVIDIA Deepen Partnership, Accelerating Chip Design with GPU-Powered AI Agents

Cadence and NVIDIA have expanded their partnership to integrate NVIDIA's GPU technologies like CUDA-X and Omniverse into Cadence's chip design software.

This collaboration shifts the heavy computing for chip simulation from traditional CPUs to powerful GPU-based 'AI factories', dramatically speeding up the design and verification process.

The initiative extends beyond chip design to 'Physical AI,' enabling the creation of digital twins for complex systems like robotics and smart factories, with Samsung and SK hynix as early adopters.

Cadence and NVIDIA have announced a major expansion of their partnership, fundamentally changing how semiconductors and complex systems are designed.

At its recent CadenceLIVE event, Cadence revealed that it is deeply embedding NVIDIA's accelerated computing technologies, including CUDA-X and Omniverse, into its core design software. This entire stack runs on Cadence's new Millennium M2000, a supercomputer powered by NVIDIA GPUs. The collaboration introduces an AI-powered 'ChipStack AI Super Agent' to automate and orchestrate highly complex, long-running design tasks, promising performance improvements of up to 80 times compared to traditional CPU-based methods.

So, what does this shift from CPU to GPU actually mean? First, it addresses a critical bottleneck in chip design. Simulating and verifying a modern chip is an incredibly intensive process that has historically relied on vast farms of CPUs, taking weeks or even months. By moving these workloads to GPUs, which are designed for parallel processing, companies can compress these cycles dramatically. This is why leading memory makers like Samsung and SK hynix are early adopters; for them, faster time-to-market for advanced chips like HBM is a significant competitive advantage.

Second, this partnership signals a broader industry pivot toward 'Physical AI.' The goal is no longer just to design a better chip in isolation. Instead, it's about simulating the entire system where that chip will live—be it a robot, a self-driving car, or an entire automated factory. By integrating with NVIDIA's Omniverse, a platform for creating physically accurate digital twins, Cadence connects the virtual world of chip design to the physical world of robotics and manufacturing. This allows engineers to test and validate how their designs will perform in real-world scenarios before anything is built.

This move also comes amid a heated competitive landscape. Rival EDA giant Synopsys recently acquired physics simulation leader Ansys and also deepened its own partnership with NVIDIA. Cadence’s focus on its AI agent and deep integration with Omniverse is a strategic response, aiming to establish a leading position in the emerging market for end-to-end, AI-driven system design.

EDA (Electronic Design Automation): A category of software tools used for designing electronic systems such as integrated circuits and printed circuit boards.
Digital Twin: A virtual model of a physical object or system. It is used to run simulations before actual devices are built and deployed.
AI Agent: An autonomous program that can perceive its environment and take actions to achieve specific goals. In this context, it automates complex chip design tasks.

Cadence and NVIDIA Deepen Partnership, Accelerating Chip Design with GPU-Powered AI Agents

Cadence and NVIDIA have expanded their partnership to integrate NVIDIA's GPU technologies like CUDA-X and Omniverse into Cadence's chip design software.

This collaboration shifts the heavy computing for chip simulation from traditional CPUs to powerful GPU-based 'AI factories', dramatically speeding up the design and verification process.

The initiative extends beyond chip design to 'Physical AI,' enabling the creation of digital twins for complex systems like robotics and smart factories, with Samsung and SK hynix as early adopters.

Cadence and NVIDIA have announced a major expansion of their partnership, fundamentally changing how semiconductors and complex systems are designed.

EDA (Electronic Design Automation): A category of software tools used for designing electronic systems such as integrated circuits and printed circuit boards.
Digital Twin: A virtual model of a physical object or system. It is used to run simulations before actual devices are built and deployed.
AI Agent: An autonomous program that can perceive its environment and take actions to achieve specific goals. In this context, it automates complex chip design tasks.