Abstract1.docx

Comparison of works on parallelization Open-MP standard was formulated as an API for writing portable, multi-threaded applications. It started as Fortran based standard, but later grew to include C and C++. The Open-MP programming models provide set of compiler programs. The result is, it describes the merge sort and Floyd’s algorithm parallely by using Open-MP to achieve good performance by reducing execution time on multi-core. For parallelizing we also use compilers. A parallelizing compiler can be created that analyses the read and write instructions and control flow modifications in programs to identify a set of dependencies between the instructions of programs. Then compiler based on generated dependencies graph rewrites and organizes the program in task oriented structure. The resulting parallel code also has the advantage of being readable and easily configured to improve further its performance. The number of parallel processors available for execution may not be known in advance. If the number of processors required to exeute the code in parallel is greater than the number of available processors, some of the modules must be executed by the same processors. Tis can be resolved by the clairvoyance framework. It is a modular environment written in Java that allows the development and testing of parallelization methods. The goals of the framework include the analysis of existing sequential progams, optimization, supporting automatic parallelization algorithms. The result of parallelization is a set of code modules that should be executed in parallel, that is, on different processors.