Enhancing the thermal performance of an electronic chipset using an innovative cooling system: Insights from machine learning models

Abstract

The thermal efficiency of an electronic chipset is investigated in this study, utilizing a novel heat sink design. A thermal energy storage system is implemented, consisting of a Phase Change Material, Heat Sink, and Metal Foam (HS-FPCM) for efficient chipset temperature control. Analyzing the metal foam's composition, PCM type, and height is part of the system's performance assessment. The HS-FPCM system is designed in three dimensions for precise evaluation, and its outputs are verified against experimental data collected under comparable operating conditions. Several Machine Learning (ML) models are built in this study to predict the HS-FPCM system outputs. The Slime Mould Algorithm (SMA) is used to optimize the ML model's hyperparameters. Based on the results, the designed ML models exhibit varying performance, with the optimized CatBoost model ranking as the best performer and the Generalized Linear Model (GLM) model as the least effective. The Time Period of a Complete Operational Cycle (TPCOC) of the electronic chipset using the aluminum and copper foam obtained to be 149 min and 154 min, respectively. Furthermore, the TPCOC values for the systems utilizing RT-35, RT-47, and RT-65 are around 149 min, 120 min, and 104 min, respectively.