基于遗传粒子群算法的三维芯片热布局优化  

Optimization of 3D Stacking Chip Thermal Layout Based on Genetic Particle Swarm Algorithm

作　　者：杨志清 潘中良[1] YANG Zhiqing;PAN Zhongliang(School of Physics and Telecommunications Engineering,South China Normal University,Guangzhou510006,China)

机构地区：[1]华南师范大学物理与电信工程学院

出　　处：《电子工艺技术》2019年第5期249-252,260,共5页Electronics Process Technology

基　　金：广州市科技计划项目(201904010107);广东省科技计划项目(2016B090918071)

摘　　要：三维叠层芯片是由多层芯片堆叠而成,其热效应与散热问题尤为突出。采用遗传粒子群算法对三维叠层芯片的热布局进行优化,研究了该芯片的功率以及个数对热布局的影响。仿真结果表明:使用遗传粒子群算法的热布局的优化精度较高,经过优化后的三维叠层芯片的温度分布更加均匀,最高温度以及温度梯度都有显著降低;三维叠层芯片的功率越高,以及堆叠的芯片个数越多,热布局优化的结果就越明显。Three-dimensional stacking chips are stacked by multi-layer chips,so their thermal effects and heat dissipation problems are particularly prominent.The thermal layout of three-dimensional stacking chips is optimized by genetic particle swarm optimization.The influence of the power and number of the stacking chips on the thermal layout is studied.The simulation results show that the genetic particle swarm optimization algorithm has high optimization accuracy.The temperature distribution of the optimized threedimensional stacked chips is more uniform,and the maximum temperature and temperature gradient decrease significantly.The higher the power of the three-dimensional stacked chips and the more stacked chips,the more obvious the thermal layout optimization results are.

关 键 词：三维堆叠芯片 热布局 热分析 遗传粒子群算法 

分 类 号：TN47[电子电信—微电子学与固体电子学]