Stochastic subspace correction methods and fault tolerance
Stochastic subspace correction methods and fault tolerance
View/Date of Publication
07.2018Publisher
Institut für Numerische Simulation (INS)Publisher Location
BonnMetadata
Show full item record
Citable Link (Handle URI) 
https://hdl.handle.net/20.500.11811/11811
Abstract
We present convergence results in expectation for stochastic subspace correction schemes and their accelerated versions to solve symmetric positive-definite variational problems, and discuss their potential for achieving fault tolerance in an unreliable compute network. We employ the standard overlapping domain decomposition algorithm for PDE discretizations to discuss the latter aspect.
Subjects
subspace correction, Schwarz iterative methods, randomization, convergence rates, fault tolerance
Classification (DDC)
510 Mathematik 518 Numerische AnalysisFirst Publication
https://ins.uni-bonn.de/publication/preprintsRelated Publications
https://doi.org/10.1090/mcom/3459Part of Series
INS Preprints ; 1809Griebel, Michael; Oswald, Peter: Stochastic subspace correction methods and fault tolerance. In: INS Preprints, 1809.
Online-Ausgabe in bonndoc: https://hdl.handle.net/20.500.11811/11811
Online-Ausgabe in bonndoc: https://hdl.handle.net/20.500.11811/11811
@unpublished{handle:20.500.11811/11811,
author = {{Michael Griebel} and {Peter Oswald}},
title = {Stochastic subspace correction methods and fault tolerance},
publisher = {Institut für Numerische Simulation (INS)},
year = 2018,
month = jul,
INS Preprints},
volume = 1809,
note = {We present convergence results in expectation for stochastic subspace correction schemes and their accelerated versions to solve symmetric positive-definite variational problems, and discuss their potential for achieving fault tolerance in an unreliable compute network. We employ the standard overlapping domain decomposition algorithm for PDE discretizations to discuss the latter aspect.},
url = {https://hdl.handle.net/20.500.11811/11811}
}
author = {{Michael Griebel} and {Peter Oswald}},
title = {Stochastic subspace correction methods and fault tolerance},
publisher = {Institut für Numerische Simulation (INS)},
year = 2018,
month = jul,
INS Preprints},
volume = 1809,
note = {We present convergence results in expectation for stochastic subspace correction schemes and their accelerated versions to solve symmetric positive-definite variational problems, and discuss their potential for achieving fault tolerance in an unreliable compute network. We employ the standard overlapping domain decomposition algorithm for PDE discretizations to discuss the latter aspect.},
url = {https://hdl.handle.net/20.500.11811/11811}
}
The following license files are associated with this item:
