Tutorial V

Modeling Electronic Systems for SI/PI: From Physics-Based Modeling to Machine Learning

Riccardo Trinchero, EMC Group, Department of Electronics and Telecommunications, Politecnico di Torino


Abstract - The modeling of complex electronic devices and systems can rely on approaches with different levels of physical knowledge, computational cost, and predictive capability. This talk provides an overview of modern modeling techniques, ranging from physics-based methods and reduced-order/macromodeling approaches to recent artificial intelligence techniques, with particular emphasis on machine learning-based surrogate models. The advantages, limitations, and trade-offs of these methodologies are discussed in the context of design optimization and uncertainty quantification. Representative applications to PCB-level structures and signal and power integrity (SI/PI) problems illustrate how machine learning complements, rather than replaces, traditional modeling techniques, enabling significant reductions in computational cost while maintaining high accuracy .

Riccardo Trinchero received the M.Sc. and Ph.D. degrees in Electronics and Communication Engineering from Politecnico di Torino, Turin, Italy, in 2011 and 2015, respectively. He is currently an Associate Professor with the EMC Group, Department of Electronics and Telecommunications, Politecnico di Torino. His research interests include the analysis of switching DC-DC converters, machine learning, and statistical simulation of circuits and systems.