Comparative Study of Multi-Threading Libraries to Fully Utilize Multi Processor/Multi Core Systems

Abstract

The development of multi-core technology has led to a great shift from sequential programming to parallel programming. This has made substantial challenges to software industries and government organizations to take full advantage of the performance intensification offered by the multi core systems. This paper aspires to compare and analyze the parallel computing ability offered by OpenMP(Open Multiprocessing), Intel Cilk Plus and MPI(Message passing Interface). Some proposals are also provided in parallel programming. The parallel programming features provided by these libraries are also studied and compared. The study is done by parallelizing problems related to Remote Sensing data processing which is large in volume and whose sequential processing is very much time consuming. This makes it pertinent to speed up the processing times by introducing parallelism in processing and for efficient utilization of multi processor multi core systems. The paper aims at exploring these libraries and studying the speed up achieved by using parallel processing and Data Parallelism paradigm.