Overview
- Cadence announced AuraStack on July 15, 2026 as an agentic “super agent” that accepts natural-language goals, plans multistep workflows and drives Cadence tools to perform PCB and advanced chip-package design.
- The platform runs on Allegro AI Studio, lets customers choose LLM providers such as OpenAI or Google, and Cadence named NVIDIA, TSMC and Schneider Electric as early collaborators with NVIDIA GPUs used to speed AI processing.
- Cadence says AuraStack continuously evaluates electrical, thermal and mechanical tradeoffs so teams share a single design context and reduce late-stage rework across board and packaging domains.
- The company claims up to 2X faster time-to-market and up to 15X task productivity and cites early customer anecdotes, including a Forvia Hella example where component placement went from about four days to roughly four minutes.
- Those performance numbers are vendor-reported and need broader, independent validation even as the launch marks a wider industry move from isolated AI assistants to integrated agentic EDA platforms that could reshape engineers’ workflows.