Analysis of Multicore CPU and GPU toward Parallelization of Total Focusing Method ultrasound reconstruction

Jason Lambert 1, 2 Antoine Pédron 1 Guillaume Gens 3 Franck Bimbard 3 Lionel Lacassagne 2 Ekaterina Iakovleva 1
1 DISC - Département Imagerie et Simulation pour le Contrôle
2 LRI - Laboratoire de Recherche en Informatique
3 Institut d′Electronique Fondamentale
Abstract : Ultrasonic imaging and reconstruction tools are commonly used to detect, identify and measure defects in different mechanical parts. Due to the complexity of the underlying physics, and due to the ever growing quantity of acquired data, computation time is becoming a limitation to the optimal inspection of a mechanical part. This article presents the performances of several implementations of a computational heavy algorithm, named Total Focusing Method, on both Graphics Processing Units (GPU) and General Purpose Processors (GPP). The scope of this study is narrowed to planar parts tested in immersion for defects. Using algorithmic simplifications and architectural optimizations, the algorithm has been drastically accelerated resulting in memory-bound implementations. On GPU, high performances can be achieved by profiting from GPU long memory transactions and from hand managed memory. Whereas on GPP, computations cost are overrun by memory access resulting in less efficient performances compared to the computing capabilities available. The following study constitutes the first step toward analyzing the target algorithm for diverse hardware in the non-destructive testing environment.
Keywords : total focusing method parallelization ultrasonic re-construction non-destructive testing gra-phic processing units general purpose processors
Type de document :
Communication dans un congrès
DASIP 2012 - Conference on Design and Architectures for Signal and Image Processing, Oct 2012, Karlsruhe, Germany. IEEE, Design and Architectures for Signal and Image Processing (DASIP), 2012 Conference on
Liste complète des métadonnées

Littérature citée [8 références]  Voir  Masquer  Télécharger
https://hal.inria.fr/hal-01093274
Contributeur : Jason Lambert <>
Soumis le : lundi 12 janvier 2015 - 10:48:08
Dernière modification le : mardi 24 avril 2018 - 13:50:57
Document(s) archivé(s) le : lundi 13 avril 2015 - 10:11:06

Fichier

dasip2012-20120514.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01093274, version 1

Collections

CEA | UMR8623 | INRIA | DRT | LIST

Citation

Jason Lambert, Antoine Pédron, Guillaume Gens, Franck Bimbard, Lionel Lacassagne, et al.. Analysis of Multicore CPU and GPU toward Parallelization of Total Focusing Method ultrasound reconstruction. DASIP 2012 - Conference on Design and Architectures for Signal and Image Processing, Oct 2012, Karlsruhe, Germany. IEEE, Design and Architectures for Signal and Image Processing (DASIP), 2012 Conference on. 〈hal-01093274〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

323

Téléchargements de fichiers

294

Contact


archive-ouverte@inria.fr