000205266 001__ 205266
000205266 005__ 20181203023750.0
000205266 0247_ $$2doi$$a10.1137/140954817
000205266 02470 $$2ISI$$a000346123200025
000205266 037__ $$aARTICLE
000205266 245__ $$aA Parallel QZ Algorithm For Distributed Memory HPC Systems
000205266 269__ $$a2014
000205266 260__ $$bSiam Publications$$c2014$$aPhiladelphia
000205266 300__ $$a24
000205266 336__ $$aJournal Articles
000205266 520__ $$aAppearing frequently in applications, generalized eigenvalue problems represent one of the core problems in numerical linear algebra. The QZ algorithm of Moler and Stewart is the most widely used algorithm for addressing such problems. Despite its importance, little attention has been paid to the parallelization of the QZ algorithm. The purpose of this work is to fill this gap. We propose a parallelization of the QZ algorithm that incorporates all modern ingredients of dense eigensolvers, such as multishift and aggressive early deflation techniques. To deal with (possibly many) infinite eigenvalues, a new parallel deflation strategy is developed. Numerical experiments for several random and application examples demonstrate the effectiveness of our algorithm on two different distributed memory HPC systems.
000205266 6531_ $$ageneralized eigenvalue problem
000205266 6531_ $$anonsymmetric QZ algorithm
000205266 6531_ $$amultishift
000205266 6531_ $$abulge chasing
000205266 6531_ $$ainfinite eigenvalues
000205266 6531_ $$aparallel algorithms
000205266 6531_ $$alevel 3 performance
000205266 6531_ $$aaggressive early deflation
000205266 700__ $$uUmea Univ, Dept Comp Sci, SE-90187 Umea, Sweden$$aAdlerborn, Bjoern
000205266 700__ $$uUmea Univ, Dept Comp Sci, SE-90187 Umea, Sweden$$aKagstroem, Bo
000205266 700__ $$0246441$$g213191$$aKressner, Daniel
000205266 773__ $$j36$$tSIAM Journal On Scientific Computing$$k5$$qC480-C503
000205266 909C0 $$xU12478$$0252494$$pANCHP
000205266 909CO $$pSB$$particle$$ooai:infoscience.tind.io:205266
000205266 917Z8 $$x213191
000205266 937__ $$aEPFL-ARTICLE-205266
000205266 973__ $$rREVIEWED$$sPUBLISHED$$aEPFL
000205266 980__ $$aARTICLE