Authors

N. Jobeir And S. Jafar

Abstract

We introduce an auto adaptive strategy enables to write a parallel algorithm adapts to the number of available resources at allocated parallel  environment to execute the parallel program. The parallel applications we are studying which are represented by data-flow graph which built dynamically during the execution. The new suggested strategy is based on coupling of a sequential algorithm and a parallel one and relies on the principle of work stealing in the tasks scheduling. We offer a study of the complexity of the adaptive algorithm and analyze its performance on processors and compare it with a performance of a classic parallel algorithm

keywords

Parallel Algorithm, Adaptive Algorithm, Data Flow Diagram, Business Theft, Tasks, Task Scheduling