Optimization Problems and algorithms

Banha University - faculty of Computers and Information

Algorithms are the key engine in numerous methods, techniques and applications. Meanwhile, The Applications’ Data is keep getting bigger and bigger exponentially. Due to these new complexities, using parallelism to enhance the performance of large-scale programs becomes more and more vital. In this paper we propose an OpenMP inspired parallel version of the Whale optimization algorithm; PWhale, it is automatically detect the number of available processors and divide the workload automatically among them to accomplish the best use of available resources. PWhale was tested using 23 benchmarks on multiple dimensions. The performance of PWhale was measured using parallel metrics such as speed, efficiency. The results of the proposed version were compared of its sequential counterpart. The comparison illustrates that the proposed version archived the same results performance while exceeding the sequential version in performance.