Carregant...
Carregant...

Vés al contingut (premeu Retorn)

Level-3 Cholesky factorization routines improve performance of many Cholesky algorithms

Autor
Gustavson, F.; Wasniewski, J.; Dongarra, J.J.; Herrero, J.; Langou, J.
Tipus d'activitat
Article en revista
Revista
ACM transactions on mathematical software
Data de publicació
2013-02
Volum
39
Número
2
Pàgina inicial
9:1
Pàgina final
9:10
DOI
https://doi.org/10.1145/2427023.2427026 Obrir en finestra nova
Resum
Four routines called DPOTF3i, i = a, b, c, d, are presented. DPOTF3i are a novel type of level-3 BLAS for use by BPF (Blocked Packed Format) Cholesky factorization and LAPACK routine DPOTRF. Performance of routines DPOTF3i are still increasing when the performance of Level-2 routine DPOTF2 of LAPACK starts decreasing. This is our main result and it implies, due to the use of larger block size nb, that DGEMM, DSYRK, and DTRSM performance also increases! The four DPOTF3i routines use simple regist...
Paraules clau
Algorithms, Performance, Lapack, Real Symmetric Matrices, Complex Hermitian Matrices, Positive Definite Matrices, Cholesky Factorization And Solution, Novel Blocked Packed Matrix Data Structures, Inplace Transposition, Cache Blocking, Blas
Grup de recerca
CAP - Grup de Computació d'Altes Prestacions

Participants

  • Gustavson, Fred G.  (autor)
  • Wasniewski, Jerzy  (autor)
  • Dongarra, Jack J.  (autor)
  • Herrero Zaragoza, José Ramón  (autor)
  • Langou, Julien  (autor)