Cadence launches AuraStack AI Super Agent to coordinate AI agents across PCB and chip-packaging design

Cadence launched AuraStack AI Super Agent on July 15, 2026, an agentic AI platform that coordinates specialized AI agents across printed circuit board (PCB) and advanced chip-packaging design. According to the company’s press release, the system runs on Cadence’s Allegro AI Studio and treats planning, implementation and multiphysics analysis — electrical, thermal and mechanical — as one coordinated flow rather than separate sequential steps. Agentic AI here means software agents that plan and carry out multi-step design tasks with limited human prompting.

Cadence says the platform delivers up to 2x faster time-to-market and a 15x productivity gain from AI-assisted automation. It cited a Forvia Hella example in which a 300-component placement task fell from about four days to about four minutes, a company-supplied figure that has not been independently verified.

The platform is accelerated by Nvidia Blackwell chips and Nvidia’s CUDA-X software libraries, a combination Cadence says yields up to 20x faster multiphysics performance. Named early users include Nvidia, TSMC, Schneider Electric and Socionext.

The label matters. A “super agent” here describes orchestration across existing design tools, not autonomous chip design. The productivity multipliers are vendor benchmarks measured on Cadence’s own examples, and results on a customer’s full board or package will depend on how much of the workflow the agents can actually own.

AuraStack is expected to become available later in 2026. Whether design teams let the agents run planning and placement end-to-end, or keep them to narrow assist tasks, will decide if the claimed time savings hold outside the demo.