Mostrar el registro sencillo del ítem
Reproducibility of parallel preconditioned conjugate gradient in hybrid programming environments
dc.contributor.author | Iakymchuk, Roman | |
dc.contributor.author | Barreda Vayá, Maria | |
dc.contributor.author | Graillat, Stef | |
dc.contributor.author | Aliaga Estellés, José Ignacio | |
dc.contributor.author | Quintana-Orti, Enrique S. | |
dc.date.accessioned | 2020-10-19T07:53:41Z | |
dc.date.available | 2020-10-19T07:53:41Z | |
dc.date.issued | 2020-06-17 | |
dc.identifier.citation | Iakymchuk R, Vayá MB, Graillat S, Aliaga JI, Quintana-Ortí ES. Reproducibility of parallel preconditioned conjugate gradient in hybrid programming environments. The International Journal of High Performance Computing Applications. 2020;34(5):502-518. doi:10.1177/1094342020932650 | ca_CA |
dc.identifier.uri | http://hdl.handle.net/10234/189995 | |
dc.description.abstract | The Preconditioned Conjugate Gradient method is often employed for the solution of linear systems of equations arising in numerical simulations of physical phenomena. While being widely used, the solver is also known for its lack of accuracy while computing the residual. In this article, we propose two algorithmic solutions that originate from the ExBLAS project to enhance the accuracy of the solver as well as to ensure its reproducibility in a hybrid MPI + OpenMP tasks programming environment. One is based on ExBLAS and preserves every bit of information until the final rounding, while the other relies upon floating-point expansions and, hence, expands the intermediate precision. Instead of converting the entire solver into its ExBLAS-related implementation, we identify those parts that violate reproducibility/non-associativity, secure them, and combine this with the sequential executions. These algorithmic strategies are reinforced with programmability suggestions to assure deterministic executions. Finally, we verify these approaches on two modern HPC systems: both versions deliver reproducible number of iterations, residuals, direct errors, and vector-solutions for the overhead of less than 37.7% on 768 cores. | ca_CA |
dc.format.extent | 17 p. | ca_CA |
dc.format.mimetype | application/pdf | ca_CA |
dc.language.iso | eng | ca_CA |
dc.publisher | Sage | ca_CA |
dc.rights | Copyright © 2020 by SAGE Publications | ca_CA |
dc.rights.uri | http://rightsstatements.org/vocab/InC/1.0/ | * |
dc.subject | preconditioned conjugate gradient | ca_CA |
dc.subject | MPI | ca_CA |
dc.subject | OpenMP tasks | ca_CA |
dc.subject | reproducibility | ca_CA |
dc.subject | accuracy | ca_CA |
dc.subject | floating-point expansion | ca_CA |
dc.subject | long accumulator | ca_CA |
dc.subject | fused multiply–add | ca_CA |
dc.title | Reproducibility of parallel preconditioned conjugate gradient in hybrid programming environments | ca_CA |
dc.type | info:eu-repo/semantics/article | ca_CA |
dc.identifier.doi | https://doi.org/10.1177/1094342020932650 | |
dc.relation.projectID | European Union’s Horizon 2020 research, innovation pro-gram under the Marie Skłodowska-Curie grant agreementvia the Robust project No. 842528 as well as the ProjectHPC-EUROPA3 (INFRAIA- 2016-1-730897), with thesupport of the H2020 EC RIA Programme ; MINECO (project TIN2017-82972-R) ; Universitat Jaume I (POSDOC-A/2017/11project) | ca_CA |
dc.rights.accessRights | info:eu-repo/semantics/openAccess | ca_CA |
dc.relation.publisherVersion | https://journals.sagepub.com/doi/full/10.1177/1094342020932650 | ca_CA |
dc.type.version | info:eu-repo/semantics/acceptedVersion | ca_CA |
Ficheros en el ítem
Este ítem aparece en la(s) siguiente(s) colección(ones)
-
ICC_Articles [424]