Design Study of Domain Decomposition Operation in Dataflow Architecture FDTD/FIT Dedicated Computer


IEICE TRANSACTIONS on Electronics   Vol.E101-C   No.1   pp.20-25
Publication Date: 2018/01/01
Online ISSN: 1745-1353
DOI: 10.1587/transele.E101.C.20
Type of Manuscript: Special Section PAPER (Special Section on Recent Progress in Electromagnetic Theory and Its Application)
Category: Electromagnetic Theory
microwave simulation,  FDTD method,  FIT,  dedicated computer,  high-performance computing,  dataflow architecture,  domain-decomposition method,  

Full Text: PDF>>
Buy this Article

To aim to achieve a high-performance computation for microwave simulations with low cost, small size machine and low energy consumption, a method of the FDTD dedicated computer has been investigated. It was shown by VHDL logical circuit simulations that the FDTD dedicated computer with a dataflow architecture has much higher performance than that of high-end PC and GPU. Then the remaining task of this work is large scale computations by the dedicated computer, since microwave simulations for only 18×18×Z grid space (Z is the number of girds for z direction) can be executed in a single FPGA at most. To treat much larger numerical model size for practical applications, this paper considers an implementation of a domain decomposition method operation of the FDTD dedicated computer in a single FPGA.