2004
|
Baldo, Lucas; Fernandes, Luiz Gustavo; Roisenberg, Paulo; Velho, Pedro; Webber, Thais Parallel PEPS Tool Performance Analysis Using Stochastic Automata Networks Inproceedings Euro-Par 2004 Parallel Processing, 10th International Euro-Par Conference (10th Euro-Par'04), pp. 214–219, Springer-Verlag, Pisa, Italy, 2004. Links @inproceedings{BAL04EUROPAR,
title = {Parallel PEPS Tool Performance Analysis Using Stochastic Automata Networks},
author = {Lucas Baldo and Luiz Gustavo Fernandes and Paulo Roisenberg and Pedro Velho and Thais Webber},
url = {https://gmap.pucrs.br/gmap/files/publications/articles/8aba43cd57e9c9da1387f812c451199a.pdf},
year = {2004},
date = {2004-05-01},
booktitle = {Euro-Par 2004 Parallel Processing, 10th International Euro-Par Conference (10th Euro-Par'04)},
pages = {214--219},
publisher = {Springer-Verlag},
address = {Pisa, Italy},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
|
2003
|
Brenner, Leonardo; Fernandes, Luiz Gustavo; Fernandes, Paulo; Sales, Afonso Performance Analysis Issues for Parallel Implementations of Propagation Algorithm Inproceedings doi 15th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD), pp. 183-191, IEEE Computer Society, Sao Paulo, Brazil, 2003. Links @inproceedings{BRE03SBACPAD,
title = {Performance Analysis Issues for Parallel Implementations of Propagation Algorithm},
author = {Leonardo Brenner and Luiz Gustavo Fernandes and Paulo Fernandes and Afonso Sales},
doi = {10.1109/CAHPC.2003.1250337},
year = {2003},
date = {2003-11-01},
booktitle = {15th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD)},
pages = {183-191},
publisher = {IEEE Computer Society},
address = {Sao Paulo, Brazil},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
|
2001
|
Denneulin, Y; Fernandes, Luiz Gustavo; Maillard, Nicolas Parallelizing a Dense Matching Region Growing Algorithm for an Image Interpolation Application. Journal Article Proceedings of the 5th International Conference on Parallel and Distributed Techniques (PDPTA), 2001. @article{PADMRGAFAIIA,
title = {Parallelizing a Dense Matching Region Growing Algorithm for an Image Interpolation Application.},
author = {Y. Denneulin and Luiz Gustavo Fernandes and Nicolas Maillard},
year = {2001},
date = {2001-01-01},
journal = {Proceedings of the 5th International Conference on Parallel and Distributed Techniques (PDPTA)},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
|
0000
|
Kolberg, Mariana; Bohlender, Gerd; Fernandes, Luiz Gustavo An efficient approach to solve very large dense linear systems with verified computing on clusters Journal Article doi Numerical Linear Algebra with Applications, 22 (2), pp. 299-316, 0000. Resumo | Links @article{gmap:KOLBERG:NLAA:14,
title = {An efficient approach to solve very large dense linear systems with verified computing on clusters},
author = {Mariana Kolberg and Gerd Bohlender and Luiz Gustavo Fernandes},
url = {https://doi.org/10.1002/nla.1950},
doi = {10.1002/nla.1950},
journal = {Numerical Linear Algebra with Applications},
volume = {22},
number = {2},
pages = {299-316},
publisher = {John Wiley & Sons},
abstract = {Automatic result verification is an important tool to guarantee that completely inaccurate results cannot be used for decisions without getting remarked during a numerical computation. Mathematical rigor provided by verified computing allows the computation of an enclosure containing the exact solution of a given problem. Particularly, the computation of linear systems can strongly benefit from this technique in terms of reliability of results. However, in order to compute an enclosure of the exact result of a linear system, more floating‐point operations are necessary, consequently increasing the execution time. In this context, parallelism appears as a good alternative to improve the solver performance. In this paper, we present an approach to solve very large dense linear systems with verified computing on clusters. This approach enabled our parallel solver to compute huge linear systems with point or interval input matrices with dimensions up to 100,000. Numerical experiments show that the new version of our parallel solver introduced in this paper provides good relative speedups and delivers a reliable enclosure of the exact results.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
Automatic result verification is an important tool to guarantee that completely inaccurate results cannot be used for decisions without getting remarked during a numerical computation. Mathematical rigor provided by verified computing allows the computation of an enclosure containing the exact solution of a given problem. Particularly, the computation of linear systems can strongly benefit from this technique in terms of reliability of results. However, in order to compute an enclosure of the exact result of a linear system, more floating‐point operations are necessary, consequently increasing the execution time. In this context, parallelism appears as a good alternative to improve the solver performance. In this paper, we present an approach to solve very large dense linear systems with verified computing on clusters. This approach enabled our parallel solver to compute huge linear systems with point or interval input matrices with dimensions up to 100,000. Numerical experiments show that the new version of our parallel solver introduced in this paper provides good relative speedups and delivers a reliable enclosure of the exact results. |