Agentic AI for Semiconductor Design and Manufacturing: From Wafer Data Workflows to Intelligent EDA Systems

Semiconductor design and manufacturing are entering a data-intensive era, with device complexity and process variability driving unprecedented computational demand. This lecture presents an emerging class of Agentic AI architectures designed to address the scale, latency, and decision-making challenges across modern EDA and fab workflows. We will examine multi-agent orchestration strategies for simulation, verification, DFM, metrology, yield learning, and root-cause discovery—supported by real industrial case studies. The session emphasizes system design choices across data models, storage, and compute pipelines, offering practical insights for circuits-and-systems innovation. Participants will learn actionable pathways to embed Agentic AI into end-to-end silicon lifecycle acceleration.