Recent Developments in Iterative Methods for Reducing Synchronization
Uloženo v:
| Název: | Recent Developments in Iterative Methods for Reducing Synchronization |
|---|---|
| Autoři: | Zou, Qinmeng, Magoulès, Frédéric |
| Přispěvatelé: | Zou, Qinmeng |
| Zdroj: | 2019 18th International Symposium on Distributed Computing and Applications for Business Engineering and Science (DCABES). :163-166 |
| Publication Status: | Preprint |
| Informace o vydavateli: | IEEE, 2019. |
| Rok vydání: | 2019 |
| Témata: | asynchronous iterations, FOS: Computer and information sciences, Computer Science - Distributed, Parallel, and Cluster Computing, pipelined Krylov subspace methods, communication-avoiding algorithms, Computer Science - Mathematical Software, [MATH] Mathematics [math], Distributed, Parallel, and Cluster Computing (cs.DC), [INFO] Computer Science [cs], 0101 mathematics, Mathematical Software (cs.MS), 01 natural sciences, s-step iterative methods |
| Popis: | On modern parallel architectures, the cost of synchronization among processors can often dominate the cost of floating-point computation. Several modifications of the existing methods have been proposed in order to keep the communication cost as low as possible. This paper aims at providing a brief overview of recent advances in parallel iterative methods for solving large-scale problems. We refer the reader to the related references for more details on the derivation, implementation, performance, and analysis of these techniques. |
| Druh dokumentu: | Article Conference object Other literature type |
| Popis souboru: | application/pdf |
| DOI: | 10.1109/dcabes48411.2019.00048 |
| DOI: | 10.48550/arxiv.1912.00816 |
| Přístupová URL adresa: | http://arxiv.org/pdf/1912.00816 http://arxiv.org/abs/1912.00816 https://hal.science/hal-02399831v2/document https://hal.science/hal-02399831v2 https://doi.org/10.1109/dcabes48411.2019.00048 https://hal.archives-ouvertes.fr/hal-02399831 https://www.arxiv-vanity.com/papers/1912.00816/ http://arxiv.org/pdf/1912.00816.pdf https://arxiv.org/abs/1912.00816 https://dblp.uni-trier.de/db/journals/corr/corr1912.html#abs-1912-00816 https://ui.adsabs.harvard.edu/abs/2019arXiv191200816Z/abstract https://arxiv.org/abs/1912.00816 |
| Rights: | IEEE Copyright arXiv Non-Exclusive Distribution |
| Přístupové číslo: | edsair.doi.dedup.....5af63843ea678c47ed80f0b10fa9946c |
| Databáze: | OpenAIRE |
Nájsť tento článok vo Web of Science | Abstrakt: | On modern parallel architectures, the cost of synchronization among processors can often dominate the cost of floating-point computation. Several modifications of the existing methods have been proposed in order to keep the communication cost as low as possible. This paper aims at providing a brief overview of recent advances in parallel iterative methods for solving large-scale problems. We refer the reader to the related references for more details on the derivation, implementation, performance, and analysis of these techniques. |
|---|---|
| DOI: | 10.1109/dcabes48411.2019.00048 |