Students from the Department of Electrical Engineering at POSTECH—Kyungjun Min, Moonkyu Choi, and Junghyun Choi (Integrated M.S.–Ph.D. students) along with six additional team members—under the supervision of Prof. Sukhyung Kang, achieved 2nd place overall in the MLCAD 2025 Contest.

The competition was jointly organized by NVIDIA and Arizona State University and was held during the Machine Learning for CAD (MLCAD) Conference in Santa Cruz, California, from September 8 to 10, 2025. The contest focused on evaluating technological capabilities in AI-based electronic design automation (EDA) for VLSI design, with 27 teams from leading universities and companies worldwide—including Samsung, IMEC, NXP, the University of Texas at Austin, Texas A&M University, and the Chinese University of Hong Kong (CUHK)—participating in the competition.

This year’s contest theme was “Physical-Aware Logic Resynthesis for Timing Optimization Using AI.” Participants were tasked with optimizing circuit timing performance through AI-driven logic resynthesis techniques.

The CSDL research team demonstrated outstanding performance, ranking first in five out of twelve benchmark circuits and achieving a total score of 713 points, securing second place overall. Fudan University ranked first with 764 points, while NXP placed third with 482 points.

This achievement highlights POSTECH’s global competitiveness in semiconductor design automation research and is expected to further strengthen Korea’s technological leadership in AI-driven system semiconductor design.