Online citations, reference lists, and bibliographies.

Fast Matrix-Free Discontinuous Galerkin Kernels On Modern Computer Architectures

M. Kronbichler, K. Kormann, I. Pasichnyk, M. Allalen
Published 2017 · Computer Science

Cite This
Download PDF
Analyze on Scholarcy
Share
This study compares the performance of high-order discontinuous Galerkin finite elements on modern hardware. The main computational kernel is the matrix-free evaluation of differential operators by sum factorization, exemplified on the symmetric interior penalty discretization of the Laplacian as a metric for a complex application code in fluid dynamics. State-of-the-art implementations of these kernels stress both arithmetics and memory transfer. The implementations of SIMD vectorization and shared-memory parallelization are detailed. Computational results are presented for dual-socket Intel Haswell CPUs at 28 cores, a 64-core Intel Knights Landing, and a 16-core IBM Power8 processor. Up to polynomial degree six, Knights Landing is approximately twice as fast as Haswell. Power8 performs similarly to Haswell, trading a higher frequency for narrower SIMD units. The performance comparison shows that simple ways to express parallelism through for loops perform better on medium and high core counts than a more elaborate task-based parallelization with dynamic scheduling according to dependency graphs, despite less memory transfer in the latter algorithm.
This paper references



This paper is referenced by
10.1002/fld.4511
Efficiency of high-performance discontinuous Galerkin spectral element methods for under-resolved turbulent incompressible flows
Niklas Fehn (2018)
10.1007/978-3-319-99654-7
Advances and New Trends in Environmental Informatics
Volker Weinberg (2018)
10.1134/S1995080219050196
Vectorization of High-performance Scientific Calculations Using AVX-512 Intruction Set
Boris Shabanov (2019)
10.1145/3322813
Multigrid for Matrix-Free High-Order Finite Element Computations on Graphics Processors
M. Kronbichler (2019)
10.1016/j.jcp.2017.09.031
On the stability of projection methods for the incompressible Navier-Stokes equations based on high-order discontinuous Galerkin discretizations
Niklas Fehn (2017)
Asynchronous Teams and Tasks in a Message Passing Environment
Benjamin Hazelwood (2019)
10.1002/fld.4683
A matrix-free high-order discontinuous Galerkin compressible Navier-Stokes solver: A performance comparison of compressible and incompressible formulations for turbulent incompressible flows
N. Fehn (2018)
10.1145/3325864
Fast Matrix-Free Evaluation of Discontinuous Galerkin Finite Element Operators
M. Kronbichler (2019)
10.1137/18M1185399
Efficient Explicit Time Stepping of High Order Discontinuous Galerkin Schemes for Waves
Svenja Schoeder (2018)
10.1007/978-3-319-99654-7_7
Efficient High-Order Discontinuous Galerkin Finite Elements with Matrix-Free Implementations
Martin Kronbichler (2018)
10.1016/j.jcp.2018.06.037
Robust and efficient discontinuous Galerkin methods for under-resolved turbulent incompressible flows
Niklas Fehn (2018)
10.1002/nme.6336
A matrix-free approach for finite-strain hyperelastic problems using geometric multigrid.
Denis Davydov (2019)
Efficient Discontinuous Galerkin Methods for Wave Propagation and Iterative Optoacoustic Image Reconstruction
S. Schoeder (2019)
10.1137/19M1276194
Enclave Tasking for Discontinuous Galerkin Methods on Dynamically Adaptive Meshes
Dominic E. Charrier (2020)
10.1016/j.jcp.2017.07.039
A high-order semi-explicit discontinuous Galerkin solver for 3D incompressible flow with application to DNS and LES of turbulent channel flow
Benjamin Krank (2017)
Semantic Scholar Logo Some data provided by SemanticScholar