Performance Evaluation of Three Dynamic Load Balancing Algorithms on Spmd Model

In this research paper, we focus on the performance of different task migration and load balancing algorithms on SPMD model based on their controlling parameters. A network of workstations has been chosen and PVM libraries have been used for implementation. Matrix multiplication has been selected as the application. Three algorithms have been investigated namely, fixed granularity, variable granularity (guided self scheduling) and one global centralized task migration algorithm. Program execution time, communication overhead and processing time have been considered as the performance measure. From the experimental results, it is found that the granularity is the controlling factor in the fixed granularity algorithm, the ratio that decreases the task size in guided self scheduling plays a vital role and that if the parameters are chosen carefully, variable task granularity and the task migration algorithms perform better than the fixed granularity.