Heat dissipation analysis and multi-objective optimization of Raspberry Pi CPU by natural convection

Abstract

With the high-speed and high-frequency operation of CPU chips, as well as miniaturization and dense assembly of integrated circuits, the heat generated by the Raspberry Pi CPU is constantly increasing. To address the urgent need for efficient Raspberry Pi CPU heat dissipation and ensure normal operation within a safe temperature range, this study first established a model to simulate the CPU's thermal physics parameters during operation using CFD simulation software. An experimental platform was then set up to validate the CFD simulation results. Based on the validated CFD model, initial heat dissipation design was carried out. On this basis, the heat dissipation system was optimized by altering the heat sink’s length, width, fin height, thickness, spacing, and the thermal grease’s thermal conductivity between the heat sink and CPU chip. A surrogate model was established using the response surface methodology, and the NSGA-II genetic algorithm was employed for multi-objective optimization of the heat dissipation system’s cooling performance. Compared to the initial solution, the chip temperature was reduced by 6.19 °C, but the heat sink mass was increased by 0.65 g. By conducting multi-objective optimization design and making reasonable choices for the parameters, a suitable and an optimum design, in terms of both economics and efficiency, is obtained. These results demonstrate that this research can provide theoretical guidance and technical support for practical production. Furthermore, the methods and findings from this study can be applied to the design of heat dissipation systems for similar applications.