A Parallel Algorithm for Multilevel k-Way Hypergraph Partitioning

Authors:
Aleksandar Trifunovic;William J. Knottenbelt
Affiliations:
Imperial College London;Imperial College London
Venue:
ISPDC '04 Proceedings of the Third International Symposium on Parallel and Distributed Computing/Third International Workshop on Algorithms, Models and Tools for Parallel Computing on Heterogeneous Networks
Year:
2004

Citing 0
Cited 7

Parallel multilevel algorithms for hypergraph partitioning

Journal of Parallel and Distributed Computing
Hypergraph Cuts & Unsupervised Representation for Image Segmentation

Fundamenta Informaticae
A general graph model for representing exact communication volume in parallel sparse matrix–vector multiplication

ISCIS'06 Proceedings of the 21st international conference on Computer and Information Sciences
Weighted adaptive neighborhood hypergraph partitioning for image segmentation

ICAPR'05 Proceedings of the Third international conference on Pattern Recognition and Image Analysis - Volume Part II
Hypergraph partitioning for faster parallel pagerank computation

EPEW'05/WS-FM'05 Proceedings of the 2005 international conference on European Performance Engineering, and Web Services and Formal Methods, international conference on Formal Techniques for Computer Systems and Business Processes
A reductive approach to hypergraph clustering: An application to image segmentation

Pattern Recognition
Early stage power management for 3D FPGAs considering hierarchical routing resources

Proceedings of the 23rd ACM international conference on Great lakes symposium on VLSI

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we present a coarse-grained parallel multi-level algorithm for the k-way hypergraph partitioning problem. The algorithm significantly improves on our previous work in terms of run time and scalability behaviour by improving processor utilisation, reducing synchronisation overhead and avoiding disk contention. The new algorithm is also generally applicable and no longer requires a particular structure of the input hypergraph to achieve a good partition quality. We present results which show that the algorithm has good scalability properties on very large hypergraphs with 脠(107) vertices and consistently outperforms the approximate partitions produced by a state-of-the-art parallel graph partitioning tool in terms of partition quality, by up to 27%.