ChipAgents at DSF 2026

Agentic AI is fundamentally changing how chips are designed. For decades, engineers have interacted with specialized tools, manually configuring workflows, analyzing results, and transferring information between design stages. In the emerging AI-native paradigm, engineers express design intent through natural language, while autonomous agents reason, collaborate, invoke tools, analyze results, and execute increasingly complex engineering workflows.

This presentation examines how agentic AI is evolving from isolated productivity features into production-ready, end-to-end flows — including coverage closure, analog behavioral model generation, formal verification, root-cause analysis, and PPA optimization. Drawing on real-world deployment results, we demonstrate how the ChipAgents agentic AI platform delivers greater engineering performance per token, higher design quality, faster time to closure, and more reliable outcomes from spec to silicon.