Title :
An Efficient Implementation of Batcher´s Odd-Even Merge Algorithm and Its Application in Parallel Sorting Schemes
Author :
Kumar, Manoj ; Hirschberg, Daniel S.
Author_Institution :
Department of Electrical Engineering, Rice University
fDate :
3/1/1983 12:00:00 AM
Abstract :
An algorithm is presented to merge two subfiles of size n/2 each, stored in the left and the right halves of a linearly connected processor array, in 3n/2 route steps and log n compare-exchange steps. This algorithm is extended to merge two horizontally adjacent subfiles of size m × n/2 each, stored in an m × n mesh-connected processor array in row-major order, in m + 2n route steps and log mn compare-exchange steps. These algorithms are faster than their counterparts proposed so far.
Keywords :
Linearly connected processor arrays; SIMD machines; mesh-connected processor arrays; odd-even merge algorithm; Algorithm design and analysis; Application software; Computer science; Hardware; Parallel processing; Registers; Sorting; Linearly connected processor arrays; SIMD machines; mesh-connected processor arrays; odd-even merge algorithm;
Journal_Title :
Computers, IEEE Transactions on
DOI :
10.1109/TC.1983.1676217