Multi-Target Vectorization With MTPS C++ Generic Library

• Main Authors: Kirschenmann, Wilfried, Plagne, Laurent, Vialle, Stéphane
• Other Authors: EDF (EDF), Algorithms for the Grid (ALGORILLE), INRIA Lorraine, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS)-Université Henri Poincaré - Nancy 1 (UHP)-Université Nancy 2-Institut National Polytechnique de Lorraine (INPL)-Centre National de la Recherche Scientifique (CNRS), SUPELEC-Campus Metz, Ecole Supérieure d'Electricité - SUPELEC (FRANCE)
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2010
• Subjects: GPU; SSE; Vectorization; c++ Template Metaprogramming; Performances; [INFO.INFO-DC]Computer Science [cs]/Distributed; Parallel; and Cluster Computing [cs.DC]; Iceland
• Online Access: https://centralesupelec.hal.science/hal-00491980; https://centralesupelec.hal.science/hal-00491980/document; https://centralesupelec.hal.science/hal-00491980/file/Supelec621.pdf

International audience This article introduces MTPS, a C++ template library dedicated at vectorizing algorithms for different target architectures. Algorithms written with MTPS benefit from optimized memory access patterns and show performances close to hardware limits, both on multicore CPU and on GPU.