Block Fusion on Dynamically Adaptive Spacetree Grids for Shallow Water Waves

Handle URI:
http://hdl.handle.net/10754/597685
Title:
Block Fusion on Dynamically Adaptive Spacetree Grids for Shallow Water Waves
Authors:
Weinzierl, Tobias; Bader, Michael; Unterweger, Kristof; Wittmann, Roland
Abstract:
© 2014 World Scientific Publishing Company. Spacetrees are a popular formalism to describe dynamically adaptive Cartesian grids. Even though they directly yield a mesh, it is often computationally reasonable to embed regular Cartesian blocks into their leaves. This promotes stencils working on homogeneous data chunks. The choice of a proper block size is sensitive. While large block sizes foster loop parallelism and vectorisation, they restrict the adaptivity's granularity and hence increase the memory footprint and lower the numerical accuracy per byte. In the present paper, we therefore use a multiscale spacetree-block coupling admitting blocks on all spacetree nodes. We propose to find sets of blocks on the finest scale throughout the simulation and to replace them by fused big blocks. Such a replacement strategy can pick up hardware characteristics, i.e. which block size yields the highest throughput, while the dynamic adaptivity of the fine grid mesh is not constrained - applications can work with fine granular blocks. We study the fusion with a state-of-the-art shallow water solver at hands of an Intel Sandy Bridge and a Xeon Phi processor where we anticipate their reaction to selected block optimisation and vectorisation.
Citation:
Weinzierl T, Bader M, Unterweger K, Wittmann R (2014) Block Fusion on Dynamically Adaptive Spacetree Grids for Shallow Water Waves. Parallel Process Lett 24: 1441006. Available: http://dx.doi.org/10.1142/S0129626414410060.
Publisher:
World Scientific Pub Co Pte Lt
Journal:
Parallel Processing Letters
KAUST Grant Number:
UK-c0020
Issue Date:
Sep-2014
DOI:
10.1142/S0129626414410060
Type:
Article
ISSN:
0129-6264; 1793-642X
Sponsors:
Tobias Weinzierl appreciates the support of the School of Engineering and Computing Sciences and in particular Tomasz Koziara at Durham University for providing
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorWeinzierl, Tobiasen
dc.contributor.authorBader, Michaelen
dc.contributor.authorUnterweger, Kristofen
dc.contributor.authorWittmann, Rolanden
dc.date.accessioned2016-02-25T12:44:23Zen
dc.date.available2016-02-25T12:44:23Zen
dc.date.issued2014-09en
dc.identifier.citationWeinzierl T, Bader M, Unterweger K, Wittmann R (2014) Block Fusion on Dynamically Adaptive Spacetree Grids for Shallow Water Waves. Parallel Process Lett 24: 1441006. Available: http://dx.doi.org/10.1142/S0129626414410060.en
dc.identifier.issn0129-6264en
dc.identifier.issn1793-642Xen
dc.identifier.doi10.1142/S0129626414410060en
dc.identifier.urihttp://hdl.handle.net/10754/597685en
dc.description.abstract© 2014 World Scientific Publishing Company. Spacetrees are a popular formalism to describe dynamically adaptive Cartesian grids. Even though they directly yield a mesh, it is often computationally reasonable to embed regular Cartesian blocks into their leaves. This promotes stencils working on homogeneous data chunks. The choice of a proper block size is sensitive. While large block sizes foster loop parallelism and vectorisation, they restrict the adaptivity's granularity and hence increase the memory footprint and lower the numerical accuracy per byte. In the present paper, we therefore use a multiscale spacetree-block coupling admitting blocks on all spacetree nodes. We propose to find sets of blocks on the finest scale throughout the simulation and to replace them by fused big blocks. Such a replacement strategy can pick up hardware characteristics, i.e. which block size yields the highest throughput, while the dynamic adaptivity of the fine grid mesh is not constrained - applications can work with fine granular blocks. We study the fusion with a state-of-the-art shallow water solver at hands of an Intel Sandy Bridge and a Xeon Phi processor where we anticipate their reaction to selected block optimisation and vectorisation.en
dc.description.sponsorshipTobias Weinzierl appreciates the support of the School of Engineering and Computing Sciences and in particular Tomasz Koziara at Durham University for providingen
dc.publisherWorld Scientific Pub Co Pte Lten
dc.subjectAdaptive Cartesian meshesen
dc.subjectBlock fusionen
dc.subjectShallow wateren
dc.subjectShared memory parallelisationen
dc.subjectSpacetreesen
dc.subjectVectorisationen
dc.titleBlock Fusion on Dynamically Adaptive Spacetree Grids for Shallow Water Wavesen
dc.typeArticleen
dc.identifier.journalParallel Processing Lettersen
dc.contributor.institutionUniversity of Durham, Durham, United Kingdomen
dc.contributor.institutionTechnische Universitat Munchen, Munich, Germanyen
kaust.grant.numberUK-c0020en
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.