Cadence VP Outlines Five Levels of AI Autonomy in Chip Design

Artificial intelligence is moving beyond its traditional role in EDA tools toward more complex tasks in chip design, according to Ziyad Hanna corporate vice president at Cadence Design Systems.

Hanna describes five levels of autonomy in chip design that mirror the automotive industrys progression toward self-driving vehicles. The levels range from basic AI-assisted design automation to fully autonomous chip creation.

AI has been used in EDA for many years for the core algorithms in tools but it is getting smarter and more optimized with the rollout of generative and agentic AI. As it evolves and improves hardware engineers are finding ways to leverage it for more complex tasks.

The highest levels of AI autonomy in chip design would allow systems to handle increasingly sophisticated design decisions with minimal human intervention. However new challenges involving traceability and explainability emerge as AI takes on more responsibility in the design process.

The semiconductor industry is pursuing these AI-driven approaches to address the growing complexity of modern chips which contain billions of transistors and require coordination across multiple engineering disciplines.