ExaGeoStat: A High Performance Unified Framework for Geostatistics on Manycore Systems

Handle URI:
http://hdl.handle.net/10754/626492
Title:
ExaGeoStat: A High Performance Unified Framework for Geostatistics on Manycore Systems
Authors:
Abdulah, Sameh; Ltaief, Hatem ( 0000-0002-6897-1095 ) ; Sun, Ying ( 0000-0001-6703-4270 ) ; Genton, Marc G. ( 0000-0001-6467-2998 ) ; Keyes, David E. ( 0000-0002-4052-7224 )
Abstract:
We present ExaGeoStat, a high performance framework for geospatial statistics in climate and environment modeling. In contrast to simulation based on partial differential equations derived from first-principles modeling, ExaGeoStat employs a statistical model based on the evaluation of the Gaussian log-likelihood function, which operates on a large dense covariance matrix. Generated by the parametrizable Matern covariance function, the resulting matrix is symmetric and positive definite. The computational tasks involved during the evaluation of the Gaussian log-likelihood function become daunting as the number n of geographical locations grows, as O(n2) storage and O(n3) operations are required. While many approximation methods have been devised from the side of statistical modeling to ameliorate these polynomial complexities, we are interested here in the complementary approach of evaluating the exact algebraic result by exploiting advances in solution algorithms and many-core computer architectures. Using state-of-the-art high performance dense linear algebra libraries associated with various leading edge parallel architectures (Intel KNLs, NVIDIA GPUs, and distributed-memory systems), ExaGeoStat raises the game for statistical applications from climate and environmental science. ExaGeoStat provides a reference evaluation of statistical parameters, with which to assess the validity of the various approaches based on approximation. The framework takes a first step in the merger of large-scale data analytics and extreme computing for geospatial statistical applications, to be followed by additional complexity reducing improvements from the solver side that can be implemented under the same interface. Thus, a single uncompromised statistical model can ultimately be executed in a wide variety of emerging exascale environments.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Publisher:
arXiv
Issue Date:
9-Aug-2017
ARXIV:
arXiv:1708.02835
Type:
Preprint
Additional Links:
http://arxiv.org/abs/1708.02835v2; http://arxiv.org/pdf/1708.02835v2
Appears in Collections:
Other/General Submission; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorAbdulah, Samehen
dc.contributor.authorLtaief, Hatemen
dc.contributor.authorSun, Yingen
dc.contributor.authorGenton, Marc G.en
dc.contributor.authorKeyes, David E.en
dc.date.accessioned2017-12-28T07:32:13Z-
dc.date.available2017-12-28T07:32:13Z-
dc.date.issued2017-08-09en
dc.identifier.urihttp://hdl.handle.net/10754/626492-
dc.description.abstractWe present ExaGeoStat, a high performance framework for geospatial statistics in climate and environment modeling. In contrast to simulation based on partial differential equations derived from first-principles modeling, ExaGeoStat employs a statistical model based on the evaluation of the Gaussian log-likelihood function, which operates on a large dense covariance matrix. Generated by the parametrizable Matern covariance function, the resulting matrix is symmetric and positive definite. The computational tasks involved during the evaluation of the Gaussian log-likelihood function become daunting as the number n of geographical locations grows, as O(n2) storage and O(n3) operations are required. While many approximation methods have been devised from the side of statistical modeling to ameliorate these polynomial complexities, we are interested here in the complementary approach of evaluating the exact algebraic result by exploiting advances in solution algorithms and many-core computer architectures. Using state-of-the-art high performance dense linear algebra libraries associated with various leading edge parallel architectures (Intel KNLs, NVIDIA GPUs, and distributed-memory systems), ExaGeoStat raises the game for statistical applications from climate and environmental science. ExaGeoStat provides a reference evaluation of statistical parameters, with which to assess the validity of the various approaches based on approximation. The framework takes a first step in the merger of large-scale data analytics and extreme computing for geospatial statistical applications, to be followed by additional complexity reducing improvements from the solver side that can be implemented under the same interface. Thus, a single uncompromised statistical model can ultimately be executed in a wide variety of emerging exascale environments.en
dc.publisherarXiven
dc.relation.urlhttp://arxiv.org/abs/1708.02835v2en
dc.relation.urlhttp://arxiv.org/pdf/1708.02835v2en
dc.rightsArchived with thanks to arXiven
dc.titleExaGeoStat: A High Performance Unified Framework for Geostatistics on Manycore Systemsen
dc.typePreprinten
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.eprint.versionPre-printen
dc.identifier.arxividarXiv:1708.02835en
kaust.authorAbdulah, Samehen
kaust.authorLtaief, Hatemen
kaust.authorSun, Yingen
kaust.authorGenton, Marc G.en
kaust.authorKeyes, David E.en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.