The BLADE paper is accepted by DAC 2026

Thanks to the contributions of all the co-authors, this work gets accepted! This work evolves from our previous TCAD paper about design automation for package power/ground planes, it proposes a bi-level Bayesian optimization framework that efficiently tackles the complex tradeoff among IR drop performance, reliability, and manufacturability. Experimental results show its great performance in synthesizing power/ground planes with much better IR drop performance, while fulfilling practical reliability and manufacturability requirements.