Multi-Target Vectorization with MTPS C++ Generic Library

Wilfried Kirschenmann (1, 2) , Laurent Plagne (1) , Stéphane Vialle (2, 3)

1 EDF - EDF
2 ALGORILLE - Algorithms for the Grid
3 IMS : Information, Multimodalité & Signal

This article introduces a C++ template library dedicated at vectorizing algorithms for different target architectures: Multi-Target Parallel Skeleton (MTPS). Skeletons describing the data structures and algorithms are provided and allow MTPS to generate a code with optimized memory access patterns for the choosen architecture. MTPS currently supports x86-64 multicore CPUs and CUDA enabled GPUs. On these architectures, performances close to hardware limits are observed.

Domains: Computer Science [cs]; Distributed, Parallel, and Cluster Computing [cs.DC]

Wilfried Kirschenmann, Laurent Plagne, Stéphane Vialle. Multi-Target Vectorization with MTPS C++ Generic Library. PARA 2010 - 10th International Conference on Applied Parallel and Scientific Computing, Jun 2010, Reykjavík, Iceland. pp.336-346, ⟨10.1007/978-3-642-28145-7_33⟩. ⟨hal-00685159⟩