Ing. Jiří Jaroš, Ph.D.

JAROŠ Jiří, VAVERKA Filip and TREEBY Bradley E. Spectral Domain Decomposition Using Local Fourier Basis: Application to Ultrasound Simulation on a Cluster of GPUs. International Journal of Supercomputing Frontiers and Innovations. South Ural State University (Chelyabinsk, Russia): Association for Computing Machinery, 2016, vol. 3, no. 3, pp. 39-54. ISSN 2313-8734. Available from: http://superfri.org/superfri/article/download/108/83
Publication language:english
Original title:Spectral Domain Decomposition Using Local Fourier Basis: Application to Ultrasound Simulation on a Cluster of GPUs
Title (cs):Dekompozice spektrálních metod pomocí lokálních Fourierových bází
Pages:39-54
Place:RU
Year:2016
URL:http://superfri.org/superfri/article/download/108/83
Journal:International Journal of Supercomputing Frontiers and Innovations, Vol. 3, No. 3, South Ural State University (Chelyabinsk, Russia), RU
ISSN:2313-8734
Files: 
+Type Name Title Size Last modified
icon2016-Jaros-SuperFRI-LocalGPU.pdf1,34 MB2017-04-04 11:09:19
^ Clear selection
With selected:
Keywords
Spectral methods, Local Fourier Basis domain decomposition, Ultrasound simulations, GPUs, HPC, Supercomputing, cancer treatment.
Annotation
The simulation of ultrasound wave propagation through biological tissue has a wide range of practical applications. However, large grid sizes are generally needed to capture the phenomena of interest. Here, a novel approach to reduce the computational complexity is presented. The model uses an accelerated k-space pseudospectral method which enables more than one hundred GPUs to be exploited to solve problems with more than 3*10^9 grid points. The classic communication bottleneck of Fourier spectral methods, all-to-all global data exchange, is overcome by the application of domain decomposition using local Fourier basis. Compared to global domain decomposition, for a grid size of 1536 x 1024 x 2048, this reduces the simulation time by a factor of 7.5 and the simulation cost by a factor of 3.8.
Abstract
The simulation of ultrasound wave propagation through biological tissue has a wide range of practical applications. However, large grid sizes are generally needed to capture the phenomena of interest. Here, a novel approach to reduce the computational complexity is presented. The model uses an accelerated k-space pseudospectral method which enables more than one hundred GPUs to be exploited to solve problems with more than 3 x 10^9 grid points. The classic communication bottleneck of Fourier spectral methods, all-to-all global data exchange, is overcome by the application of domain decomposition using local Fourier basis. Compared to global domain decomposition, for a grid size of 1536 x 1024 x 2048, this reduces the simulation time by a factor of 7.5 and the simulation cost by a factor of 3.8.
BibTeX:
@ARTICLE{
   author = {Ji{\v{r}}{\'{i}} Jaro{\v{s}} and Filip Vaverka and E.
	Bradley Treeby},
   title = {Spectral Domain Decomposition Using Local Fourier Basis:
	Application to Ultrasound Simulation on a Cluster of GPUs},
   pages = {39--54},
   journal = {International Journal of Supercomputing Frontiers and
	Innovations},
   volume = {3},
   number = {3},
   year = {2016},
   ISSN = {2313-8734},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=11149}
}

Your IPv4 address: 54.81.108.205
Switch to IPv6 connection

DNSSEC [dnssec]