Acceleration of Le Bail fitting method on parallel platforms

Mařík, Ondřej, and Šimeček, Ivan. "Acceleration of Le Bail fitting method on parallel platforms." Programs and Algorithms of Numerical Mathematics. Prague: Institute of Mathematics AS CR, 2015. 136-141. <http://eudml.org/doc/269935>.

@inProceedings{Mařík2015,
abstract = {Le Bail fitting method is procedure used in the applied crystallography mainly during the crystal structure determination. As in many other applications, there is a need for a great performance and short execution time. In this paper, we describe utilization of parallel computing for mathematical operations used in Le Bail fitting. We present an algorithm implementing this method with highlighted possible approaches to its aforementioned parallelization. Then, we propose a sample parallel version using the OpenMP API and its performance results on the real multithreaded system. Further potential for the massive parallelization is also discussed.},
author = {Mařík, Ondřej, Šimeček, Ivan},
booktitle = {Programs and Algorithms of Numerical Mathematics},
keywords = {Le Bail fitting; applied crystallography; nonlinear least squares; parallel computing; OpenMP},
location = {Prague},
pages = {136-141},
publisher = {Institute of Mathematics AS CR},
title = {Acceleration of Le Bail fitting method on parallel platforms},
url = {http://eudml.org/doc/269935},
year = {2015},
}

TY - CLSWK
AU - Mařík, Ondřej
AU - Šimeček, Ivan
TI - Acceleration of Le Bail fitting method on parallel platforms
T2 - Programs and Algorithms of Numerical Mathematics
PY - 2015
CY - Prague
PB - Institute of Mathematics AS CR
SP - 136
EP - 141
AB - Le Bail fitting method is procedure used in the applied crystallography mainly during the crystal structure determination. As in many other applications, there is a need for a great performance and short execution time. In this paper, we describe utilization of parallel computing for mathematical operations used in Le Bail fitting. We present an algorithm implementing this method with highlighted possible approaches to its aforementioned parallelization. Then, we propose a sample parallel version using the OpenMP API and its performance results on the real multithreaded system. Further potential for the massive parallelization is also discussed.
KW - Le Bail fitting; applied crystallography; nonlinear least squares; parallel computing; OpenMP
UR - http://eudml.org/doc/269935
ER -