Automatic performance tuning of parallel and accelerated seismic imaging kernels

Handle URI:
http://hdl.handle.net/10754/564848
Title:
Automatic performance tuning of parallel and accelerated seismic imaging kernels
Authors:
Haberdar, Hakan; Siddiqui, Shahzeb; Feki, Saber
Abstract:
With the increased complexity and diversity of mainstream high performance computing systems, significant effort is required to tune parallel applications in order to achieve the best possible performance for each particular platform. This task becomes more and more challenging and requiring a larger set of skills. Automatic performance tuning is becoming a must for optimizing applications such as Reverse Time Migration (RTM) widely used in seismic imaging for oil and gas exploration. An empirical search based auto-tuning approach is applied to the MPI communication operations of the parallel isotropic and tilted transverse isotropic kernels. The application of auto-tuning using the Abstract Data and Communication Library improved the performance of the MPI communications as well as developer productivity by providing a higher level of abstraction. Keeping productivity in mind, we opted toward pragma based programming for accelerated computation on latest accelerated architectures such as GPUs using the fairly new OpenACC standard. The same auto-tuning approach is also applied to the OpenACC accelerated seismic code for optimizing the compute intensive kernel of the Reverse Time Migration application. The application of such technique resulted in an improved performance of the original code and its ability to adapt to different execution environments.
KAUST Department:
Core Labs
Publisher:
EAGE Publications
Journal:
EAGE Workshop on High Performance Computing for Upstream
Conference/Event name:
EAGE Workshop on High Performance Computing for Upstream 2014
Issue Date:
2014
DOI:
10.3997/2214-4609.20141941
Type:
Conference Paper
ISBN:
9781634391672
Appears in Collections:
Conference Papers

Full metadata record

DC FieldValue Language
dc.contributor.authorHaberdar, Hakanen
dc.contributor.authorSiddiqui, Shahzeben
dc.contributor.authorFeki, Saberen
dc.date.accessioned2015-08-04T07:23:00Zen
dc.date.available2015-08-04T07:23:00Zen
dc.date.issued2014en
dc.identifier.isbn9781634391672en
dc.identifier.doi10.3997/2214-4609.20141941en
dc.identifier.urihttp://hdl.handle.net/10754/564848en
dc.description.abstractWith the increased complexity and diversity of mainstream high performance computing systems, significant effort is required to tune parallel applications in order to achieve the best possible performance for each particular platform. This task becomes more and more challenging and requiring a larger set of skills. Automatic performance tuning is becoming a must for optimizing applications such as Reverse Time Migration (RTM) widely used in seismic imaging for oil and gas exploration. An empirical search based auto-tuning approach is applied to the MPI communication operations of the parallel isotropic and tilted transverse isotropic kernels. The application of auto-tuning using the Abstract Data and Communication Library improved the performance of the MPI communications as well as developer productivity by providing a higher level of abstraction. Keeping productivity in mind, we opted toward pragma based programming for accelerated computation on latest accelerated architectures such as GPUs using the fairly new OpenACC standard. The same auto-tuning approach is also applied to the OpenACC accelerated seismic code for optimizing the compute intensive kernel of the Reverse Time Migration application. The application of such technique resulted in an improved performance of the original code and its ability to adapt to different execution environments.en
dc.publisherEAGE Publicationsen
dc.titleAutomatic performance tuning of parallel and accelerated seismic imaging kernelsen
dc.typeConference Paperen
dc.contributor.departmentCore Labsen
dc.identifier.journalEAGE Workshop on High Performance Computing for Upstreamen
dc.conference.date7 September 2014 through 10 September 2014en
dc.conference.nameEAGE Workshop on High Performance Computing for Upstream 2014en
dc.contributor.institutionUniversity of Houston, United Statesen
kaust.authorFeki, Saberen
kaust.authorSiddiqui, Shahzeben
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.