a Only synchronization is termination detection. All shared data are read-only. b Edge sharing with neighbors. Compute maximum change between iterations. c Fork/join parallelism. No data sharing. d Decreasing work. Every iteration disseminate values to all. e Variable amount of work.
A comparison of parallel implementation of explicit DG and central difference method
✍ Scribed by Ekevid, Torbjörn ;Wiberg, Nils-Erik
- Publisher
- John Wiley and Sons
- Year
- 2002
- Tongue
- English
- Weight
- 206 KB
- Volume
- 18
- Category
- Article
- ISSN
- 1069-8299
- DOI
- 10.1002/cnm.515
No coin nor oath required. For personal study only.
📜 SIMILAR VOLUMES
In this paper a parallel algorithm is created for solving electromagnetic wave scattering problems. The algorithm based on the finite difference time-domain (FDTD) representation of the Maxwell's equations written in 3L Parallel Fortran for use on a transputer array. The problem simulated consists
A coarse-grain, parallel approach to direct Hartree-Fock calculations is presented and discussed. The suggested scheme allows for a near asymptotic speedup involving a very low parallelization overhead without compromising the vector performance of vector-parallel architectures. A shared-memory MIMD
The finite difference time-domain (FDTD) method is a well known numerical technique that has been used to solve electromagnetic boundary value problems. However, the method requires large computational resources to solve a problem, restricting its use on sequential computers to small problems. This