Optimization Techniques and Performance Evaluation of a Multithreaded Multi-core Architecture Using OpenMP

Optimization techniques are the crucial steps in parallelizing the programs for multi-core architectures. These Multi-Core architectures have become more popular due to improvement in performance, power concerns, thermal dissipations and more efficient simultaneous processing of multi tasks. There are wide variety of optimization techniques, but there is no history notify about the order in which these techniques to be applied for a program to obtain maximum performance. This paper mainly focuses on analysis of various program optimization techniques for multi-Core architectures. Finally this paper shows how the sequential code can be parallelized using OpenMP programming environment and will be explaining the interest findings with V-Tune analyzer.

[1]  Barbara M. Chapman,et al.  Performance modeling of communication and computation in hybrid MPI and OpenMP applications , 2006, 12th International Conference on Parallel and Distributed Systems - (ICPADS'06).

[2]  Rudolf Eigenmann,et al.  Programming Distributed Memory Sytems Using OpenMP , 2007, 2007 IEEE International Parallel and Distributed Processing Symposium.

[3]  Dean M. Tullsen,et al.  Proximity-aware directory-based coherence for multi-core processor architectures , 2007, SPAA '07.

[4]  Kleanthis Psarris,et al.  Program analysis techniques for transforming programs for parallel execution , 2002, Parallel Comput..

[5]  Glen I. Magee.  Code Optimization Techniques , 2000 .