Implementing parallel sorting algorithms on a linear array of transputers

Authors:
W. P. Goodwin;S. K. Das
Affiliations:
E-Systems, Inc., Greenville Division, Greenville, Texas;Department of Computer Science, University of North Texas, Denton, Texas
Venue:
Proceedings of the 1989 ACM/IEEE conference on Supercomputing
Year:
1989

Citing 3
Cited 1

Transputer reference manual

Transputer reference manual
Communicating sequential processes

Communications of the ACM
Parallel Sorting Algorithms

Parallel Sorting Algorithms

Practical aspects and experiences: Performance of parallel spanning tree algorithms on linear arrays of transputers and unix systems

Parallel Computing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents experimental results on a linear array of transputers for three parallel sorting algorithms, odd-even transposition sort, weavesort, and pipelined-merge sort. Though a significant amount of message communication is usually required for implementing sorting algorithms on any linear array of processors, in our experiments it is observed that the actual time spent for this is not a major overhead. This is because of the high communication bandwidth provided by the transputer system. Each of the three algorithms are described in a high-level parallel language. The number of message communications and merging complexities of these algorithms are analyzed and compared. We plot performance parameters, including parallel time complexities and resulting speedups, as functions of the number of processors used and the number of elements to be sorted.