International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 98 - Issue 13 |
Published: July 2014 |
Authors: Jae-Hyeon Parq, Erik Sevre, Sang-Mook Lee |
![]() |
Jae-Hyeon Parq, Erik Sevre, Sang-Mook Lee . Effects of Easy Hybrid Parallelization with CUDA for OpenMX. International Journal of Computer Applications. 98, 13 (July 2014), 20-27. DOI=10.5120/17244-7580
@article{ 10.5120/17244-7580, author = { Jae-Hyeon Parq,Erik Sevre,Sang-Mook Lee }, title = { Effects of Easy Hybrid Parallelization with CUDA for OpenMX }, journal = { International Journal of Computer Applications }, year = { 2014 }, volume = { 98 }, number = { 13 }, pages = { 20-27 }, doi = { 10.5120/17244-7580 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2014 %A Jae-Hyeon Parq %A Erik Sevre %A Sang-Mook Lee %T Effects of Easy Hybrid Parallelization with CUDA for OpenMX%T %J International Journal of Computer Applications %V 98 %N 13 %P 20-27 %R 10.5120/17244-7580 %I Foundation of Computer Science (FCS), NY, USA
A MPI-friendly density functional theory (DFT) source code was modified within hybrid parallelization including CUDA. The objective is to find out how simple conversions within the hybrid parallelization with mid-range GPUs affect DFT code not originally suitable to CUDA. Several rules of hybrid parallelization for numerical-atomic-orbital (NAO) DFT codes were settled. The test was performed on a magnetite material system with OpenMX code by utilizing a hardware system containing 2 Xeon E5606 CPUs and 2 Quadro 4000 GPUs. 3-way hybrid routines obtained a speedup of 7. 55 while 2-way hybrid speedup by 10. 94. GPUs with CUDA complement the efficiency of OpenMP and compensate CPUs' excessive competition within MPI.