Abstract Near-optimal CPU efficiency for the basic operations of the Newton-Krylov-BILU iterations for the discontinuous Galerkin method (DGM) can be obtained by exploiting its data locality, as it allows to rewrite these operations in terms of BLAS/LAPACK operations on contiguous vectors and matrices of considerable size. Further significant enhancements are obtained by using single precision pre-conditioners and by exploiting data alignment to improve cache efficiency. The underlying data structures and operations are explained, followed by the comparison of the obtained floating point operation (FLOP) efficiency to the theoretical optimum.
Exploiting Data Locality in the DGM Discretisation for Optimal Efficiency
2010-01-01
13 pages
Article/Chapter (Book)
Electronic Resource
English
Double Precision , Discontinuous Galerkin Method , Single Precision , GMRES Iteration , Convective Interface Engineering , Engineering Fluid Dynamics , Astronomy, Astrophysics and Cosmology , Classical Continuum Physics , Aerospace Technology and Astronautics , Appl.Mathematics/Computational Methods of Engineering , Numerical and Computational Physics
Using Finite Element Discretisation for Spacecraft Thermal Analysis
SAE Technical Papers | 1995
|Implementing a Hybrid Space Discretisation within an Agent Based Evacuation Model
British Library Conference Proceedings | 2011
|EVALUATION OF NAUTICAL MINUTE DISCRETISATION FOR CONTROL OF CONTINENTAL ENROUTE AIRSPACE
British Library Conference Proceedings | 2010
|