首页> 外文会议>Global Symposium on Millimeter-Waves and Terahertz >Electromagnetic-Circuital-Thermal Multiphysics Simulation Based on DGTD and FETD Method with Higher-Order Basis Functions
【24h】

Electromagnetic-Circuital-Thermal Multiphysics Simulation Based on DGTD and FETD Method with Higher-Order Basis Functions

机译:基于DGTD和FETD方法的电磁 - 电路 - 热多体仿真,具有高阶基函数

获取原文

摘要

With the continuous improvement of the technological level of integrated circuits (ICs), the operating frequency and integration density are becoming higher and higher, which brings huge challenges to the simulation of integrated circuits. On one hand, the wave effect becomes noticeable as the working frequency increases. For instance, the electromagnetic coupling between the interconnects and package will lead to the signal integrity problem. In this situation, the simulation accuracy cannot be guaranteed if only the classical circuit theory is adopted. Therefore, full-wave electromagnetic simulation method must be included to implement electromagnetic-circuital cosimulation. On the other hand, high integration level will not only raise the density of the devices and interconnects, but also increase the power density, resulting in difficulties in thermal management. As a result, the mutual effects between EM and thermal must also be taken into account. To sum up, the electromagnetic-circuital-thermal multiphysics simulation becomes necessary and significant in the design of ICs. In this paper, an electromagnetic-thermal co-simulation method based on discontinuous Galerkin time-domain (DGTD) method and finite-element time-domain (FETD) method with higher order basis functions is proposed. The DGTD method is utilized for EM simulation [2], while the FETD method is adopted for thermal simulation [3]–[5]. The higher order hierachical basis function is utilized to decrease the mesh density and reduce the number of spatial unknowns. The large-scale parallel computing technique based on message passing interface (MPI) is employed to accelerate both the DGTD and FETD algorithm. Numerical examples are computed on distributed clusters to validate the accuracy and efficiency of the proposed method.
机译:随着集成电路(IC)技术水平的不断改进,工作频率和集成密度变得越来越高,为集成电路的模拟带来了巨大挑战。一方面,随着工作频率的增加,波动效应变得明显。例如,互连和包之间的电磁耦合将导致信号完整性问题。在这种情况下,如果仅采用经典电路理论,则无法保证模拟精度。因此,必须包括全波电磁仿真方法以实现电磁 - 电路算法。另一方面,高集成电平不仅会提高器件的密度和互连,而且还增加了功率密度,导致热管理中的困难。结果,还必须考虑EM和热量之间的相互影响。总而言之,在IC的设计中,电磁 - 电路 - 热多体多体模拟变得必要和重要。本文提出了一种基于不连续的Galerkin时域(DGTD)方法和具有更高阶基函数的有限元时间域(FETD)方法的电磁热共计仿真方法。 DGTD方法用于EM仿真[2],而采用FETD方法进行热仿真[3] - [5]。使用更高阶的格尔查氏函数来降低网状密度并减少空间未知数的数量。基于消息传递接口(MPI)的大规模并行计算技术用于加速DGTD和FETD算法。在分布式集群上计算数值示例以验证所提出的方法的准确性和效率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号