Machine learning-based solution for thermo-mechanical analysis of MMIC packaging

Abstract

Abstract Thermo-mechanical analysis of monolithic microwave integrated circuit (MMIC) packaging is essential to guarantee the reliability of radio frequency/microwave applications. However, a method for fast and accurate analysis of MMIC packaging structures has not been developed. Here, we demonstrate a machine learning (ML)-based solution for thermo-mechanical analysis of MMIC packaging. This ML-based solution analyzes temperature and thermal stresses considering 13 design parameters categorized into material properties, geometric characteristics, and thermal boundary conditions. Finite element simulation with the Monte Carlo method is utilized to prepare 40,000 data samples for supervised learning and validation of the ML solution, and a laser-assisted thermal experiment verifies the accuracy of the simulation. After data preparation, regression tree ensemble and artificial neural network (ANN) learning models are investigated. The results indicate that the ANN models accurately predict the temperature and thermal stresses, showing a 1.69 % minimum error. Finally, the developed ML solution is deployed as a web application format for facile approaches. We believe that this study will provide a guideline for developing ML-based solutions in chip packaging design technology.