Computing least squares condition numbers on hybrid multicore/GPU systems
Universite Paris XI – Paris Sud
hal-00947204, (14 February 2014)
@techreport{baboulin:hal-00947204,
hal_id={hal-00947204},
url={http://hal.inria.fr/hal-00947204},
title={Computing least squares condition numbers on hybrid multicore/GPU systems},
author={Baboulin, Marc and Dongarra, Jack and Lacroix, R{‘e}mi},
keywords={linear least squares; condition number; statistical condition estimation; variance-covariance; GPU computing; MAGMA library},
language={Anglais},
affiliation={Laboratoire de Recherche en Informatique – LRI , POSTALE – INRIA Saclay – Ile de France , Innovative Computing Laboratory – ICL , ALPINES – INRIA Paris-Rocquencourt},
type={Rapport de recherche},
institution={INRIA},
number={RR-8479},
year={2014},
month={Feb},
pdf={http://hal.inria.fr/hal-00947204/PDF/RR-8479.pdf}
}
This paper presents an efficient computation for least squares conditioning or estimates of it. We propose performance results using new routines on top of the multicore-GPU library MAGMA. This set of routines is based on an efficient computation of the variance-covariance matrix for which, to our knowledge, there is no implementation in current public domain libraries LAPACK and ScaLAPACK.
February 21, 2014 by hgpu
