Towards Fast Reverse Time Migration Kernels using Multi-threaded Wavefront Diamond Tiling

Handle URI:
http://hdl.handle.net/10754/578819
Title:
Towards Fast Reverse Time Migration Kernels using Multi-threaded Wavefront Diamond Tiling
Authors:
Malas, T.; Hager, G.; Ltaief, Hatem ( 0000-0002-6897-1095 ) ; Keyes, David E. ( 0000-0002-4052-7224 )
Abstract:
Today’s high-end multicore systems are characterized by a deep memory hierarchy, i.e., several levels of local and shared caches, with limited size and bandwidth per core. The ever-increasing gap between the processor and memory speed will further exacerbate the problem and has lead the scientific community to revisit numerical software implementations to better suit the underlying memory subsystem for performance (data reuse) as well as energy efficiency (data locality). The authors propose a novel multi-threaded wavefront diamond blocking (MWD) implementation in the context of stencil computations, which represents the core operation for seismic imaging in oil industry. The stencil diamond formulation introduces temporal blocking for high data reuse in the upper cache levels. The wavefront optimization technique ensures data locality by allowing multiple threads to share common adjacent point stencil. Therefore, MWD is able to take up the aforementioned challenges by alleviating the cache size limitation and releasing pressure from the memory bandwidth. Performance comparisons are shown against the optimized 25-point stencil standard seismic imaging scheme using spatial and temporal blocking and demonstrate the effectiveness of MWD.
KAUST Department:
Extreme Computing Research Center
Publisher:
EAGE Publications
Journal:
Second EAGE Workshop on High Performance Computing for Upstream
Conference/Event name:
Second EAGE Workshop on High Performance Computing for Upstream
Issue Date:
13-Sep-2015
DOI:
10.3997/2214-4609.201414025
Type:
Conference Paper
Additional Links:
http://www.earthdoc.org/publication/publicationdetails/?publication=82624
Appears in Collections:
Conference Papers; Extreme Computing Research Center

Full metadata record

DC FieldValue Language
dc.contributor.authorMalas, T.en
dc.contributor.authorHager, G.en
dc.contributor.authorLtaief, Hatemen
dc.contributor.authorKeyes, David E.en
dc.date.accessioned2015-09-28T13:51:09Zen
dc.date.available2015-09-28T13:51:09Zen
dc.date.issued2015-09-13en
dc.identifier.doi10.3997/2214-4609.201414025en
dc.identifier.urihttp://hdl.handle.net/10754/578819en
dc.description.abstractToday’s high-end multicore systems are characterized by a deep memory hierarchy, i.e., several levels of local and shared caches, with limited size and bandwidth per core. The ever-increasing gap between the processor and memory speed will further exacerbate the problem and has lead the scientific community to revisit numerical software implementations to better suit the underlying memory subsystem for performance (data reuse) as well as energy efficiency (data locality). The authors propose a novel multi-threaded wavefront diamond blocking (MWD) implementation in the context of stencil computations, which represents the core operation for seismic imaging in oil industry. The stencil diamond formulation introduces temporal blocking for high data reuse in the upper cache levels. The wavefront optimization technique ensures data locality by allowing multiple threads to share common adjacent point stencil. Therefore, MWD is able to take up the aforementioned challenges by alleviating the cache size limitation and releasing pressure from the memory bandwidth. Performance comparisons are shown against the optimized 25-point stencil standard seismic imaging scheme using spatial and temporal blocking and demonstrate the effectiveness of MWD.en
dc.publisherEAGE Publicationsen
dc.relation.urlhttp://www.earthdoc.org/publication/publicationdetails/?publication=82624en
dc.titleTowards Fast Reverse Time Migration Kernels using Multi-threaded Wavefront Diamond Tilingen
dc.typeConference Paperen
dc.contributor.departmentExtreme Computing Research Centeren
dc.identifier.journalSecond EAGE Workshop on High Performance Computing for Upstreamen
dc.conference.date13-16 September 2015en
dc.conference.nameSecond EAGE Workshop on High Performance Computing for Upstreamen
dc.conference.locationDubai, UAEen
dc.eprint.versionPost-printen
dc.contributor.institutionErlangen Regional Computing Centeren
kaust.authorKeyes, David E.en
kaust.authorMalas, T.en
kaust.authorLtaief, Hatemen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.