Δημοσιεύσεις

Project Acronym: ScaleSciComp
Title: Scale Scientific Computations
Affiliation: democritus university of thrace
Pi: George Gravvanis
Research Field: mathematics and computer sciences

Parallel multi-projection preconditioned methods based on semi-aggregation techniques
by Byron E. Moutafis, Christos K. Filelis-Papadopoulos and George A. Gravvanis
Abstract:
Abstract A parallel preconditioned iterative method in conjunction with an additive domain decomposition method based on semi-aggregation techniques for general sparse linear systems is presented. The proposed preconditioning scheme constitutes an algebraic non-overlapping domain decomposition method in fine-coarse (semi-coarse) subdomains, which is based only on the graph of the corresponding coefficient matrix and aggregation techniques. The proposed scheme is a hybrid algorithm designed for distributed systems with multicore nodes. Numerical results concerning the convergence behavior and the parallel performance of the proposed method are given along with discussions.
Reference:
Parallel multi-projection preconditioned methods based on semi-aggregation techniques (Byron E. Moutafis, Christos K. Filelis-Papadopoulos and George A. Gravvanis), In Journal of Computational Science, volume 22, 2017.
Bibtex Entry:
@article{MOUTAFIS201745,
 title = {Parallel multi-projection preconditioned methods based on semi-aggregation techniques},
 journal = {Journal of Computational Science},
 volume = {22},
 number = {Supplement C},
 pages = {45 - 53},
 year = {2017},
 bibyear = {2017},
 issn = {1877-7503},
 doi = {https://doi.org/10.1016/j.jocs.2017.08.020},
 url = {http://www.sciencedirect.com/science/article/pii/S1877750317304647},
 author = {Byron E. Moutafis and Christos K. Filelis-Papadopoulos and George A. Gravvanis},
 abstract = {Abstract A parallel preconditioned iterative method in conjunction with an additive domain decomposition method based on semi-aggregation techniques for general sparse linear systems is presented. The proposed preconditioning scheme constitutes an algebraic non-overlapping domain decomposition method in fine-coarse (semi-coarse) subdomains, which is based only on the graph of the corresponding coefficient matrix and aggregation techniques. The proposed scheme is a hybrid algorithm designed for distributed systems with multicore nodes. Numerical results concerning the convergence behavior and the parallel performance of the proposed method are given along with discussions.},
}