Reliability Modeling Approaches: Physics or AI/ML

[]This presentation provides a high-level description of CALCE-UMD activities in reliability physics of microelectronic systems, starting with a brief history and continuing on to recent trends in multiscale modeling of the reliability of advanced microelectronic packaging. The discussion includes specific focus on the importance of considering material microstructure in predictive reliability physics modeling; and explores the role of reliability physics in the context of AI/ML* approaches for reliability modeling.
In Topic 1, we will examine three examples where microstructure-sensitive modeling can provide important insights into material behavior: (i) organic interposers/substrates that are based on fabric-reinforced composites; (ii) solder alloys with heterogeneous multiscale microstructure; (iii) sintered silver materials with agglomerated nanoporous microstructure. In Topic 2, we will qualitatively explore the interplay between reliability physics and AI/ML in influencing both epistemic as well as aleatory uncertainties in reliability predictions.
*AI/ML: Artificial Intelligence / Machine Learning
Speaker(s): Abhijit Dasgupta,
Bldg: ARMS 3115, Purdue University, West Lafayette, Indiana, United States, Virtual: https://events.vtools.ieee.org/m/554090