Designing parallel algorithms for parallel system

Abstract EN

Modularity, scalability, simultaneousness and modularity are four attractive qualities related to parallel algorithms presented in this research.

Simultaneousness alludes to capacity for performing numerous activities at the same time; this is basic in the case where the program needs to be executed on numerous processors.

Scalability shows strength to expanding the number of processors and is similarly useful, as processor include seem liable to develop in the majority of general conditions.

Locality implies a high proportion of neighborhood memory accessing to remote memory accessing (correspondence); this is the way to optimal efficiency on multi-computer models.

Modularity the decomposing of complicated elements into less complex parts is a fundamental part of programming building, in parallel processing and also traditional computing.

The multi-computer parallel computers show, and the task/channel programming model presented in this Task is going to be utilized as a part of resulting dialog of parallel algorithm structure, examination, and usage.

The multi-computer comprises of at least a single von Neumann PCs associated by organized interconnections.

It’s a basic and practical device demonstrate that gives a premise to the outline of scalable and convenient parallel projects.

A programming structure in light of assignments and channels improves programming multi-computers via giving reflections which enable discussing correspondence, locality and concurrency in machine-free system, also via giving a base to specific development of parallel projects.