Title :
FDTD simulations with 1011 unknowns using AVX and SSD on a consumer PC
Author :
Simon, Winfried ; Lauer, Andreas ; Wien, Andreas
Author_Institution :
IMST GmbH, Kamp-Lintfort, Germany
Abstract :
This paper presents how modern consumer PCs and server workstations can be used very efficiently for large scale FDTD simulations. The development of the computer systems in the last years going from single core CPUs to multi core CPUs with improved calculation extensions like AVX (Advanced Vector Extension) enables new possibilities for implementing fast FDTD algorithms. A new XPU calculation technique (software accelerated computing on modern CPUs) which speeds up FDTD simulations is explained and verified by several large and complex application examples. This XPU technique easily surpasses the speed and size of calculation on GPU cards by using consumer PCs.
Keywords :
finite difference time-domain analysis; graphics processing units; mathematics computing; multiprocessing systems; workstation clusters; AVX; CPU; FDTD simulations; GPU cards; SSD; XPU calculation; advanced vector extension; consumer PC; server workstations; software accelerated computing; Computational modeling; Finite difference methods; Graphics processing units; Memory management; Power dividers; Time domain analysis; Workstations;
Conference_Titel :
Antennas and Propagation Society International Symposium (APSURSI), 2012 IEEE
Conference_Location :
Chicago, IL
Print_ISBN :
978-1-4673-0461-0
DOI :
10.1109/APS.2012.6348422