Introducing Direct Mapping Sorters For Parallel Sorting Algorithms
Pushpendra Kumar1, Priyanka Tyagi2, Smriti Joshi3
1Pushpendra Kumar, Meerut Institute of Engineering and Technology, Meerut.
2Priyanka Tyagi, Meerut Institute of Engineering and Technology, Meerut.
3Smriti Joshi, Meerut Institute of Engineering and Technology, Meerut.
Manuscript received on November 28, 2012. | Revised Manuscript received on December 16, 2012. | Manuscript published on December 30, 2012. | PP: 145-148 | Volume-2, Issue-2, December 2012. | Retrieval Number: B0886112212 /2012©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Sorting is one of the most basic problems of computer science and has been discussed continuously since the evolution of computer science. Several algorithms have been devised and applied and the work is still unfinished. For the parallel computing sorting is of same relevance as for sequential and very primitive problem domain too. Grain size is very important aspect of any parallel algorithm and is decisive in term of complexity. For the sorting problems minimum unit for sorting is two elements, since we apply a swap operation if required, and the two elements are sorted. This is considered to be the single step operation. In this paper we will increase primitive unit to four elements and four elements will be sorted in a single step. By applying this technique we can improve the performance of many parallel algorithms.
Keywords: Parallel Sorting, Bitonic, Shear Sort, Direct Mapping.