KAIST student receives best paper award at international symposium

KAIST graduate student Lee Jung-hyun won the Best Student Paper - Honorable Mention at the 2026 IEEE International Symposium on Electromagnetic Compatibility, Signal & Power Integrity held in Dallas, Texas from August 3-7. The award recognized Lee's paper titled "Physics-Aware Tensor Learning for Data-Efficient Multi-Port Power Distribution Network Analysis."

This research introduced a novel tensor-based method to handle and analyze large volumes of power distribution network (PDN) data in complex semiconductor packages like high-bandwidth memory (HBM) and chiplet systems. The key innovation involved utilizing a mathematical approach called Tucker decomposition to identify and store essential patterns from the data.

This technique achieved an impressive compression ratio of approximately 695 to 1 in a test involving a 256-port PDN, while maintaining the desired level of accuracy. Lee's interest in tensors stemmed from his early exposure to this concept.