A scalable RBF-FD method for atmospheric flow

JOURNAL OF COMPUTATIONAL PHYSICS

卷:298

Article

Tillenius, Martin¹  Larsson, Elisabeth¹  Lehto, Erik²  Flyer, Natasha³ 

[1] Uppsala Univ, Dept Informat Technol, SE-75105 Uppsala, Sweden

[2] KTH Royal Inst Technol, Dept Math, SE-10044 Stockholm, Sweden

[3] Natl Ctr Atmospher Res, Boulder, CO 80307 USA

关键词: Shallow water;    Scattered node;    Task parallel;    Distributed memory;    Multicore;    Radial basis function;    RBF-FD;   

DOI  :  10.1016/j.jcp.2015.06.003

来源: Elsevier

【 摘 要 】

Radial basis function-generated finite difference (RBF-FD) methods have recently been proposed as very interesting for global scale geophysical simulations, and have been shown to outperform established pseudo-spectral and discontinuous Galerkin methods for shallow water test problems. In order to be competitive for very large scale simulations, the RBF-FD methods needs to be efficiently implemented for modern multicore based computer architectures. This is a challenging assignment, because the main computational operations are unstructured sparse matrix-vector multiplications, which in general scale poorly on multicore computers due to bandwidth limitations. However, with the task parallel implementation described here we achieve 60-100% of theoretical speedup within a shared memory node, and 80-100% of linear speedup across nodes. We present results for global shallow water benchmark problems with a 30 km resolution. (C) 2015 Elsevier Inc. All rights reserved.

【 授权许可】

Free