Cargando…

Travel Software using GPU Hardware

Travel is the main multi-particle tracking code being used at CERN for the beam dynamics calculations through hadron and ion linear accelerators. It uses two routines for the calculation of space charge forces, namely, rings of charges and point-to-point. This report presents the studies to improve...

Descripción completa

Detalles Bibliográficos
Autores principales: Szalwinski, Chris M, Belli, Eleonora, Dimov, Veliko Atanasov
Lenguaje:eng
Publicado: 2015
Materias:
Acceso en línea:http://cds.cern.ch/record/2063005
_version_ 1780948571762196480
author Szalwinski, Chris M
Belli, Eleonora
Dimov, Veliko Atanasov
author_facet Szalwinski, Chris M
Belli, Eleonora
Dimov, Veliko Atanasov
author_sort Szalwinski, Chris M
collection CERN
description Travel is the main multi-particle tracking code being used at CERN for the beam dynamics calculations through hadron and ion linear accelerators. It uses two routines for the calculation of space charge forces, namely, rings of charges and point-to-point. This report presents the studies to improve the performance of Travel using GPU hardware. The studies showed that the performance of Travel with the point-to-point simulations of space-charge effects can be speeded up at least 72 times using current GPU hardware. Simple recompilation of the source code using an Intel compiler can improve performance at least 4 times without GPU support. The limited memory of the GPU is the bottleneck. Two algorithms were investigated on this point: repeated computation and tiling. The repeating computation algorithm is simpler and is the currently recommended solution. The tiling algorithm was more complicated and degraded performance. Both build and test instructions for the parallelized version of the software are included.
id cern-2063005
institution Organización Europea para la Investigación Nuclear
language eng
publishDate 2015
record_format invenio
spelling cern-20630052019-09-30T06:29:59Zhttp://cds.cern.ch/record/2063005engSzalwinski, Chris MBelli, EleonoraDimov, Veliko AtanasovTravel Software using GPU HardwareAccelerators and Storage RingsTravel is the main multi-particle tracking code being used at CERN for the beam dynamics calculations through hadron and ion linear accelerators. It uses two routines for the calculation of space charge forces, namely, rings of charges and point-to-point. This report presents the studies to improve the performance of Travel using GPU hardware. The studies showed that the performance of Travel with the point-to-point simulations of space-charge effects can be speeded up at least 72 times using current GPU hardware. Simple recompilation of the source code using an Intel compiler can improve performance at least 4 times without GPU support. The limited memory of the GPU is the bottleneck. Two algorithms were investigated on this point: repeated computation and tiling. The repeating computation algorithm is simpler and is the currently recommended solution. The tiling algorithm was more complicated and degraded performance. Both build and test instructions for the parallelized version of the software are included. CERN-ACC-NOTE-2015-0034oai:cds.cern.ch:20630052015-10-28
spellingShingle Accelerators and Storage Rings
Szalwinski, Chris M
Belli, Eleonora
Dimov, Veliko Atanasov
Travel Software using GPU Hardware
title Travel Software using GPU Hardware
title_full Travel Software using GPU Hardware
title_fullStr Travel Software using GPU Hardware
title_full_unstemmed Travel Software using GPU Hardware
title_short Travel Software using GPU Hardware
title_sort travel software using gpu hardware
topic Accelerators and Storage Rings
url http://cds.cern.ch/record/2063005
work_keys_str_mv AT szalwinskichrism travelsoftwareusinggpuhardware
AT bellieleonora travelsoftwareusinggpuhardware
AT dimovvelikoatanasov travelsoftwareusinggpuhardware