Improved Performance for Nodal Spectral Element Operators

Authors:
Uwe Fladrich;Jörg Stiller;Wolfgang E. Nagel
Affiliations:
CENTER FOR INFORMATION SERVICES AND HIGH PERFORMANCECOMPUTING (ZIH), TECHNISCHE UNIVERSITÄT DRESDEN, 01062 DRESDEN, GERMANY;INSTITUTE FOR AEROSPACE ENGINEERING (ILR), TECHNISCHEUNIVERSITÄT DRESDEN, 01062 DRESDEN, GERMANY;CENTER FOR INFORMATION SERVICES AND HIGH PERFORMANCECOMPUTING (ZIH), TECHNISCHE UNIVERSITÄT DRESDEN, 01062 DRESDEN, GERMANY
Venue:
International Journal of High Performance Computing Applications
Year:
2008

Citing 7
Cited 0

the LINPACK benchmark: an explanation

Proceedings of the 1st International Conference on Supercomputing
Spectral methods on triangles and other domains

Journal of Scientific Computing
A generalized diagonal mass matrix spectral element method for non-quadrilateral elements

Proceedings of the fourth international conference on Spectral and high order methods (ICOSAHOM 1998)
A pseudo-spectral scheme for the incompressible Navier-Stokes equations using unstructured nodal elements

Journal of Computational Physics
Stable Spectral Methods on Tetrahedral Elements

SIAM Journal on Scientific Computing
Nodal high-order methods on unstructured grids

Journal of Computational Physics
Factorization Techniques for Nodal Spectral Elements in Curved Domains

SIAM Journal on Scientific Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

The computational performance of the nodal spectral element method (SEM) for tetrahedral grids is evaluated in the context of a high-performance computing platform. The elemental SEM operator is accelerated by the symmetry-based factorization technique of Hesthaven and Teng (SIAM J. Sci. Comp. 21, p. 2352, 2000), which results in a reduced number of floating point operations and memory accesses. However, performance evaluation reveals that a naive implementation of the algorithm causes a severe degradation of computational efficiency. Two algorithmic modifications are proposed which regain (and partly exceed) the efficiency of the original, non-factorized operator and thus recover the asymptotic 9/5 speedup due to factorization.