Performance Optimization of Parallel Programming

Abstract — A well organized parallel application can accomplish better performance over sequential execution on existing and forthcoming parallel PC architecture. This paper depicts the issues that can limit the performance of parallel programs that restricts them from providing the expected performance improvement compared to sequential programming, test assessment of certain parallel application with sequential programs. Before depicting the exploratory assessment, this paper depicts a few systems pertinent of parallel programming. The assessment of parallel applications has been carried out by exploratory result assessment and execution estimations. Proper utilization of all available cores and resources, in addition to allocating balanced amount of load among all computing units, can lead to improved performance