OpenMP-based parallel algorithms for solving kronecker descriptors

A.M. De Lima; T. Webber; M.A.S. Netto; R.M. Czekster; C.A.F. De Rose; P. Fernandes

doi:10.1109/SBAC-PADW.2010.13

OpenMP-based parallel algorithms for solving kronecker descriptors

A.M. De Lima, T. Webber, M.A.S. Netto, R.M. Czekster, C.A.F. De Rose, P. Fernandes

Research output: Chapter in Book/Published conference output › Conference publication

Abstract

Numerical analysis of Markovian models is relevant for performance evaluation and probabilistic analysis of systems' behavior from several fields such as Bioinformatics, Economics, and Engineering. These models can be represented in a compact fashion using Kronecker algebra. The Vector-Descriptor Product is the key operation to obtain stationary solutions of Kronecker-based descriptors. Due to its complexity, the numerical algorithms are usually CPU intensive, requiring alternatives such as data partitioning in order to produce results in less time. This paper proposes three OpenMP-based parallel implementations for solving descriptors to be deployed on shared-memory machines. We evaluated the implementations in a multi-core machine and obtained a speed-up near to eight when using eight cores with Intel Hyper-Threading technology.

Original language	English
Title of host publication	Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA
Publisher	IEEE
ISBN (Print)	978-1-4244-8877-3
DOIs	https://doi.org/10.1109/SBAC-PADW.2010.13
Publication status	Published - 29 Nov 2010

Access to Document

10.1109/SBAC-PADW.2010.13

Cite this

De Lima, A. M., Webber, T., Netto, M. A. S., Czekster, R. M., De Rose, C. A. F., & Fernandes, P. (2010). OpenMP-based parallel algorithms for solving kronecker descriptors. In Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA IEEE. https://doi.org/10.1109/SBAC-PADW.2010.13

@inproceedings{2f20cff964784686b7e4a7a101617356,

title = "OpenMP-based parallel algorithms for solving kronecker descriptors",

abstract = "Numerical analysis of Markovian models is relevant for performance evaluation and probabilistic analysis of systems' behavior from several fields such as Bioinformatics, Economics, and Engineering. These models can be represented in a compact fashion using Kronecker algebra. The Vector-Descriptor Product is the key operation to obtain stationary solutions of Kronecker-based descriptors. Due to its complexity, the numerical algorithms are usually CPU intensive, requiring alternatives such as data partitioning in order to produce results in less time. This paper proposes three OpenMP-based parallel implementations for solving descriptors to be deployed on shared-memory machines. We evaluated the implementations in a multi-core machine and obtained a speed-up near to eight when using eight cores with Intel Hyper-Threading technology.",

author = "{De Lima}, A.M. and T. Webber and M.A.S. Netto and R.M. Czekster and {De Rose}, C.A.F. and P. Fernandes",

year = "2010",

month = nov,

day = "29",

doi = "10.1109/SBAC-PADW.2010.13",

language = "English",

isbn = "978-1-4244-8877-3",

booktitle = "Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA",

publisher = "IEEE",

address = "United States",

}

De Lima, AM, Webber, T, Netto, MAS, Czekster, RM, De Rose, CAF & Fernandes, P 2010, OpenMP-based parallel algorithms for solving kronecker descriptors. in Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA. IEEE. https://doi.org/10.1109/SBAC-PADW.2010.13

OpenMP-based parallel algorithms for solving kronecker descriptors. / De Lima, A.M.; Webber, T.; Netto, M.A.S. et al.
Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA. IEEE, 2010.

Research output: Chapter in Book/Published conference output › Conference publication

TY - GEN

T1 - OpenMP-based parallel algorithms for solving kronecker descriptors

AU - De Lima, A.M.

AU - Webber, T.

AU - Netto, M.A.S.

AU - Czekster, R.M.

AU - De Rose, C.A.F.

AU - Fernandes, P.

PY - 2010/11/29

Y1 - 2010/11/29

N2 - Numerical analysis of Markovian models is relevant for performance evaluation and probabilistic analysis of systems' behavior from several fields such as Bioinformatics, Economics, and Engineering. These models can be represented in a compact fashion using Kronecker algebra. The Vector-Descriptor Product is the key operation to obtain stationary solutions of Kronecker-based descriptors. Due to its complexity, the numerical algorithms are usually CPU intensive, requiring alternatives such as data partitioning in order to produce results in less time. This paper proposes three OpenMP-based parallel implementations for solving descriptors to be deployed on shared-memory machines. We evaluated the implementations in a multi-core machine and obtained a speed-up near to eight when using eight cores with Intel Hyper-Threading technology.

AB - Numerical analysis of Markovian models is relevant for performance evaluation and probabilistic analysis of systems' behavior from several fields such as Bioinformatics, Economics, and Engineering. These models can be represented in a compact fashion using Kronecker algebra. The Vector-Descriptor Product is the key operation to obtain stationary solutions of Kronecker-based descriptors. Due to its complexity, the numerical algorithms are usually CPU intensive, requiring alternatives such as data partitioning in order to produce results in less time. This paper proposes three OpenMP-based parallel implementations for solving descriptors to be deployed on shared-memory machines. We evaluated the implementations in a multi-core machine and obtained a speed-up near to eight when using eight cores with Intel Hyper-Threading technology.

UR - http://www.scopus.com/inward/record.url?eid=2-s2.0-78650834338&partnerID=MN8TOARS

UR - https://ieeexplore.ieee.org/document/5644919

U2 - 10.1109/SBAC-PADW.2010.13

DO - 10.1109/SBAC-PADW.2010.13

M3 - Conference publication

SN - 978-1-4244-8877-3

BT - Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA

PB - IEEE

ER -

De Lima AM, Webber T, Netto MAS, Czekster RM, De Rose CAF, Fernandes P. OpenMP-based parallel algorithms for solving kronecker descriptors. In Proceedings - 22nd International Symposium on Computer Architecture and High Performance Computing Workshops, SBAC-PADW 2010, 1st Workshop on Applications for Multi and Many Core Architectures, WAMMCA. IEEE. 2010 doi: 10.1109/SBAC-PADW.2010.13

OpenMP-based parallel algorithms for solving kronecker descriptors

Abstract

Access to Document

Other files and links

Fingerprint

Cite this