GPU algorithms for Efficient Exascale Discretizations


Abdelfattah A., Barra V., Beams N., Bleile R., Brown J., Camier J., ...More

Parallel Computing, vol.108, 2021 (Peer-Reviewed Journal) identifier identifier

  • Publication Type: Article / Article
  • Volume: 108
  • Publication Date: 2021
  • Doi Number: 10.1016/j.parco.2021.102841
  • Journal Name: Parallel Computing
  • Journal Indexes: Science Citation Index Expanded, Scopus, Academic Search Premier, Applied Science & Technology Source, Computer & Applied Sciences, INSPEC, zbMATH
  • Keywords: High-performance computing, GPU acceleration, High-order discretizations, Finite element methods, Exascale applications, FINITE-ELEMENT-METHOD, ORDER, PERFORMANCE, ACCELERATION, INTEGRATION, OPENACC

Abstract

© 2021 Elsevier B.V.In this paper we describe the research and development activities in the Center for Efficient Exascale Discretization within the US Exascale Computing Project, targeting state-of-the-art high-order finite-element algorithms for high-order applications on GPU-accelerated platforms. We discuss the GPU developments in several components of the CEED software stack, including the libCEED, MAGMA, MFEM, libParanumal, and Nek projects. We report performance and capability improvements in several CEED-enabled applications on both NVIDIA and AMD GPU systems.