Cadence And Google Collaborate To Expand AI-Driven EDAs On Google Cloud

Cadence has announced a collaboration with Google to integrate its ChipStack AI Super Agent with Gemini models on Google Cloud, aiming to scale AI-driven chip design and verification workflows through a cloud-based platform. The initiative focuses on advancing agent-based design automation by combining large language model (LLM) capabilities with electronic design automation (EDA) tools, which is used to develop chip-level circuitry.

Cadence ChipStack AI Super Agent

The ChipStack AI Super Agent – now available in Google Cloud Marketplace – is designed to incorporate agent-based reasoning with Cadence’s existing EDA software, supporting tasks such as digital design, testbench development, verification planning, regression management, and automated debugging. The integration with Gemini models is intended to enhance the system’s ability to process design workflows and generate outputs with improved consistency and efficiency.

Cadence reports that the platform can deliver up to 10x productivity improvements across selected design and verification processes under specific conditions – a key component of the system is its Mental Model technology, which enables coordinated interaction between AI-generated outputs and Cadence’s native design tools, with the aim of improving the accuracy and usability of generated content.

The collaboration also uses Google Cloud infrastructure to provide scalable compute resources required for AI processing and EDA workloads, supporting deployment through a cloud-native environment. The combined solution is designed to allow deployment through a simplified setup process, enabling organizations to run agent-based chip design and verification workflows on Google Cloud.

Pokdepinion: NVIDIA touted similar productivity boosts in its internal chip design tools with AI before, and this is the stuff where LLM is particularly good at.