Parallelizing the merge sorting network algorithm on a multi-core computer using Go and Cilk

Authors:
Peiyi Tang;Doug Serfass
Affiliations:
University of Arkansas at Little Rock, Little Rock, AR;University of Arkansas at Little Rock, Little Rock, AR
Venue:
Proceedings of the 49th Annual Southeast Regional Conference
Year:
2011

Citing 5
Cited 0

Cilk: an efficient multithreaded runtime system

PPOPP '95 Proceedings of the fifth ACM SIGPLAN symposium on Principles and practice of parallel programming
How to Sort N Items Using a Sorting Network of Fixed I/O Size

IEEE Transactions on Parallel and Distributed Systems
Introduction to Algorithms

Introduction to Algorithms
Parallel Programming in C with MPI and OpenMP

Parallel Programming in C with MPI and OpenMP
Sorting Networks of Logarithmic Depth, Further Simplified

Algorithmica

Quantified Score

Hi-index	0.00

Visualization

Abstract

We create a scalable parallel algorithm based on parallelizing the merge sorting network algorithm. We implemented our scalable parallel algorithm using two modern shared-memory parallel programming languages, Go [1] and Cilk-5 [2]. We also compare Go and Cilk speedup and performance. Our experimental results show that our parallel algorithm is scalable and provides near linear speedup for all large problem sizes. We compare Go and Cilk scheduling and synchronization efficiency as well as the overall performance of our parallel code.