OpenMP versus MPI for PDE solvers based on regular sparse numerical operators.
Markus NordénSverker HolmgrenMichael ThunéPublished in: Future Gener. Comput. Syst. (2006)
Keyphrases
- shared memory
- high performance computing
- parallel programming
- partial differential equations
- message passing interface
- parallel computing
- message passing
- numerical algorithms
- parallel algorithm
- numerically stable
- massively parallel
- numerical methods
- distributed memory
- high order
- sparse data
- morphological operators
- multiscale
- parallel implementation
- compressive sensing
- scale spaces
- high dimensional
- multi core processors
- multi processor
- general purpose
- finite difference
- parallelization strategy
- computing systems
- anisotropic diffusion
- sparse representation
- parallel computation
- computing resources
- parallel execution
- differential operators
- grid computing
- sensitivity analysis
- image denoising
- finite difference method
- level set