Attributes | Values |
---|
rdf:type
| |
Description
| - Používáme paralelní přímý řešič na bázi Schurových doplňků pro řešení velkých řídkých lineárních systémů, vzniklých z metody konečných prvků. Doménová dekompozice se provádí pomocí dělení grafu a vede na řídké podmatice. Tyto podmatice se ukladájí v paměti a faktorizují se obálkovou metodou. Ale napřed se proměnné v každé podmatici přečíslují Sloanovým algoritmem, aby se minimalizovaly kvality: paměťové nároky a čas faktorizace. Tyto kvality jsou obyčejně nevyvážené. V tomto příspěvku popisujeme výsledky integrace Quality Balancing (QB) heuristiky, která vyvažuje kvality, a modifikace Sloanova algoritmu, nazývanou Boundary Sloan algoritmus (BSA). Diskutujeme problémy v chování QB heuristiky a vyhodnocujeme implementaci 3 variant Sloanova algoritmu s QB heuristikou. (cs)
- We use a parallel direct solver based on the Schur complement method for solving large sparse linear systems arising from the finite element method. A domain decomposition of a problem is performed using a graph partitioning. It results in sparse submatrices. An envelope method is used to store these submatrices in the memory and to factorize them. Prior to the solution, the variables of each submatrix are reordered by the Sloan algorithm to minimize qualities: memory requirements and the factorization time. These qualities are usually disbalanced. In this paper, we describe results of integrating our recently developed Quality Balancing (QB) heuristics and our modification of the Sloan algorithm, called Boundary Sloan algorithm (BSA), to balance the qualities. We discuss the issues of behaviour of the QB heuristics and evaluate the results of implementation of 3 variants of the Sloan algorithm within the QB heuristics.
- We use a parallel direct solver based on the Schur complement method for solving large sparse linear systems arising from the finite element method. A domain decomposition of a problem is performed using a graph partitioning. It results in sparse submatrices. An envelope method is used to store these submatrices in the memory and to factorize them. Prior to the solution, the variables of each submatrix are reordered by the Sloan algorithm to minimize qualities: memory requirements and the factorization time. These qualities are usually disbalanced. In this paper, we describe results of integrating our recently developed Quality Balancing (QB) heuristics and our modification of the Sloan algorithm, called Boundary Sloan algorithm (BSA), to balance the qualities. We discuss the issues of behaviour of the QB heuristics and evaluate the results of implementation of 3 variants of the Sloan algorithm within the QB heuristics. (en)
|
Title
| - Quality Balancing Heuristics with Three Variants of Sloan Algorithm
- Heuristika pro vyvažování kvalit se třemi variantami hraničního Sloanova algoritmu (cs)
- Quality Balancing Heuristics with Three Variants of Sloan Algorithm (en)
|
skos:prefLabel
| - Quality Balancing Heuristics with Three Variants of Sloan Algorithm
- Heuristika pro vyvažování kvalit se třemi variantami hraničního Sloanova algoritmu (cs)
- Quality Balancing Heuristics with Three Variants of Sloan Algorithm (en)
|
skos:notation
| - RIV/68407700:21230/05:03107453!RIV06-MSM-21230___
|
http://linked.open.../vavai/riv/strany
| |
http://linked.open...avai/riv/aktivita
| |
http://linked.open...avai/riv/aktivity
| - P(IBS3086102), Z(MSM6840770014)
|
http://linked.open...vai/riv/dodaniDat
| |
http://linked.open...aciTvurceVysledku
| |
http://linked.open.../riv/druhVysledku
| |
http://linked.open...iv/duvernostUdaju
| |
http://linked.open...titaPredkladatele
| |
http://linked.open...dnocenehoVysledku
| |
http://linked.open...ai/riv/idVysledku
| - RIV/68407700:21230/05:03107453
|
http://linked.open...riv/jazykVysledku
| |
http://linked.open.../riv/klicovaSlova
| - Sloan algorithm; domain docomposition; load balancing; parallel solver (en)
|
http://linked.open.../riv/klicoveSlovo
| |
http://linked.open...ontrolniKodProRIV
| |
http://linked.open...v/mistoKonaniAkce
| |
http://linked.open...i/riv/mistoVydani
| |
http://linked.open...i/riv/nazevZdroje
| - Proceedings of the IASTED International Conference on PARALLEL AND DISTRIBUTED COMPUTING AND NETWORKS PDCN 2005
|
http://linked.open...in/vavai/riv/obor
| |
http://linked.open...ichTvurcuVysledku
| |
http://linked.open...cetTvurcuVysledku
| |
http://linked.open...vavai/riv/projekt
| |
http://linked.open...UplatneniVysledku
| |
http://linked.open...iv/tvurceVysledku
| - Tvrdík, Pavel
- Medek, Ondřej
|
http://linked.open...vavai/riv/typAkce
| |
http://linked.open.../riv/zahajeniAkce
| |
http://linked.open...n/vavai/riv/zamer
| |
number of pages
| |
http://purl.org/ne...btex#hasPublisher
| |
https://schema.org/isbn
| |
http://localhost/t...ganizacniJednotka
| |
is http://linked.open...avai/riv/vysledek
of | |