Geodesic
A parallel preconditioner based on the approximation of an inverse matrix by power series for solving sparse linear systems on graphics processors
Numerical methods and programming, Tome 20 (2019) no. 4, pp. 444-456.

Voir la notice de l'article provenant de la source Math-Net.Ru

The applicability of the AIPS method approximating an inverse matrix using Neumann series is considered in the framework of the CPR two stage preconditioner. A parallel CUDA-oriented algorithm is proposed for solving linear systems with tridiagonal matrices consisting of independent blocks of different sizes. It is shown that the implementation of the proposed algorithm can be more than twice the speed of the similar functions from the cuSPARSE library. Experimental evaluation of the BiCGStab method with the CPR-AIPS preconditioner on modern GPUs, including a hybrid computing system with 4 GPU NVIDIA Tesla V100, is performed. Numerical experiments show an adequate scalability of this preconditioner as well as the possibility (compared to the CPR-AMG) to accelerate the solution of linear systems being typical for the reservoir modeling problems.
Keywords: CUDA, graphics processors, iterative methods, parallel computing, preconditioners, tridiagonal systems.
Mots-clés : sparse matrices 
@article{VMP_2019_20_4_a8,
     author = {A. V. Yuldashev and N. V. Repin and V. V. Spele},
     title = {A parallel preconditioner based on the approximation of an inverse matrix by power series for solving sparse linear systems on graphics processors},
     journal = {Numerical methods and programming},
     pages = {444--456},
     publisher = {mathdoc},
     volume = {20},
     number = {4},
     year = {2019},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/VMP_2019_20_4_a8/}
}
TY  - JOUR
AU  - A. V. Yuldashev
AU  - N. V. Repin
AU  - V. V. Spele
TI  - A parallel preconditioner based on the approximation of an inverse matrix by power series for solving sparse linear systems on graphics processors
JO  - Numerical methods and programming
PY  - 2019
SP  - 444
EP  - 456
VL  - 20
IS  - 4
PB  - mathdoc
UR  - http://geodesic.mathdoc.fr/item/VMP_2019_20_4_a8/
LA  - ru
ID  - VMP_2019_20_4_a8
ER  - 
%0 Journal Article
%A A. V. Yuldashev
%A N. V. Repin
%A V. V. Spele
%T A parallel preconditioner based on the approximation of an inverse matrix by power series for solving sparse linear systems on graphics processors
%J Numerical methods and programming
%D 2019
%P 444-456
%V 20
%N 4
%I mathdoc
%U http://geodesic.mathdoc.fr/item/VMP_2019_20_4_a8/
%G ru
%F VMP_2019_20_4_a8
A. V. Yuldashev; N. V. Repin; V. V. Spele. A parallel preconditioner based on the approximation of an inverse matrix by power series for solving sparse linear systems on graphics processors. Numerical methods and programming, Tome 20 (2019) no. 4, pp. 444-456. http://geodesic.mathdoc.fr/item/VMP_2019_20_4_a8/