KAIST Terah Lab to Host Workshop on HBM Design Automation Using AI Agents

KAIST Terah Lab, led by Kim Joungho, Professor at the School of Electrical Engineering at KAIST, announced on July 2 that it will host a workshop titled "OpenClaw AI Agent-driven High Bandwidth Memory (HBM) Design and Research Work Automation" on July 3 from 8 AM to 12 PM.

Kim Joungho, Professor at KAIST

This workshop is organized to share cases in which generative artificial intelligence (AI) and AI agent technologies have been applied to semiconductor design and research and development (R&D) fields. In particular, it will introduce methods for increasing design automation and research productivity in the high bandwidth memory (HBM) and advanced semiconductor packaging design sectors through the use of AI agents.

The workshop will present various use cases centered on the OpenClaw AI agent platform, which has been developed and is actively in use by KAIST Terah Lab. These include design automation, simulation automation, document management, and research support. Graduate student researchers from Terah Lab, including both master’s and doctoral students, will participate directly in the presentations.

The first session, under the theme of "Research Assistant," will introduce the OpenClaw AI agent platform and its Model Context Protocol (MCP)-based utilization methods. Topics will include the AI agent for package design and analysis using KiCAD MCP, the AI agent for semiconductor power delivery network simulation, the AI agent for eye diagram dataset construction and equalizer optimization, and cases of simulation automation using PyAEDT-based Ansys EM tools.

The second session, themed "Daily Assistant," will cover AI agent use cases that support researchers’ daily work tasks, such as AI-based portfolio management using OpenClaw, an automated stock trading system integrated with securities firm APIs, automatic collection of paper keywords and generation of relevance graphs, and the creation of presentation drafts based on Google Calendar events.

This workshop is significant in that it shares examples of implementing AI as agents that perform actual work, rather than as simple Q&A tools, to automate the entire semiconductor design and research and development process. Practical methods to simultaneously enhance design efficiency and research productivity in the HBM and advanced packaging fields are expected to be presented.

Seo Hae-seok, a doctoral student at KAIST Terah Lab, stated, "Through this workshop, which showcases Terah Lab’s experience in building AI agent-based research environments, we hope to see the adoption of AI-driven R&D automation spread to other research institutions in the semiconductor industry."

KAIST Terah Lab plans to release a video of the workshop presentations through the lab’s official website in the future.

This content was produced with the assistance of AI translation services.