摘要
提出一种全新的时域有限元运算加速方法——应用图形处理器(GPU)实现对时域有限元运算的提速。在GPU上通过OpenGL编程完成时域有限元运算,从而利用GPU的并行处理功能达到提高运算速度的目的。文中介绍了编程实现过程,并分析讨论了GPU提速过程中处理数据量不同具有不同表现的现象,进而得出随着处理数据量的增加,GPU相对于CPU的处理速度将会更快。文中以时域有限元的辐射算例验证了该算法的正确性和有效性,为时域有限元运算提速提供了一条有效途径。
In this paper,a new accelerated method for Time-Domain Finite Element Method（TD-FEM） is presented——the speed of time-domain finite element method is accelerated by Graphics Processing Unit（GPU）.The programming methods have been displayed in this paper,and the phenomenon that different data quantity has different effect is analyzed and discussed.As a result,if the amount of calculation is increasing,comparing with a Central Processing Unit（CPU）,the speed of Graphics Processing Unit（GPU） is more faster.The correctness and effectiveness of this method are proved by a radiation sample with TD-FEM.A new way to accelerate the computation speed is provided for TD-FEM.
出处
《微波学报》
CSCD
北大核心
2010年第5期50-52,67,共4页
Journal of Microwaves
基金
成都信息工程学院科研基金(KYTZ201019)
