Three Versions of Clique Search Parallelization
Abstract
In our paper we present three parallel versions for maximum clique finding algorithms. The parallelization algorithm by quasi coloring was proposed by S. Szabo. In our paper we present the actual implementation of the algorithm and actual results measured on a large scale supercomputer up to 512 cores. We also implemented the proposed tuning of the algorithm and present here the measurements and results. Apart for the implementation we propose another method of tuning the original algorithm, which is based on Las Vegas randomization method. In our paper we compare the measured results of the three versions of the algorithm.
Full Text: PDF
Abstract
In our paper we present three parallel versions for maximum clique finding algorithms. The parallelization algorithm by quasi coloring was proposed by S. Szabo. In our paper we present the actual implementation of the algorithm and actual results measured on a large scale supercomputer up to 512 cores. We also implemented the proposed tuning of the algorithm and present here the measurements and results. Apart for the implementation we propose another method of tuning the original algorithm, which is based on Las Vegas randomization method. In our paper we compare the measured results of the three versions of the algorithm.
Full Text: PDF
Browse Journals
Journal Policies
Information
Useful Links
- Call for Papers
- Submit Your Paper
- Publish in Your Native Language
- Subscribe the Journal
- Frequently Asked Questions
- Contact the Executive Editor
- Recommend this Journal to Librarian
- View the Current Issue
- View the Previous Issues
- Recommend this Journal to Friends
- Recommend a Special Issue
- Comment on the Journal
- Publish the Conference Proceedings
Latest Activities
Resources
Visiting Status
 Today |
68 |
| Yesterday |
215 |
| This Month |
332 |
| Last Month |
6586 |
 All Days |
1467045 |
 Online |
12 |
