Loading [a11y]/accessibility-menu.js
Ant Brood Clustering on Intel Xeon Multi-core: Challenges and Strategies | IEEE Conference Publication | IEEE Xplore

Ant Brood Clustering on Intel Xeon Multi-core: Challenges and Strategies


Abstract:

Swarm intelligence algorithms such as ant colony optimization (ACO) and particle swarm optimization (PSO) are computationally intensive. Ant Brood Clustering (ABC) algori...Show More

Abstract:

Swarm intelligence algorithms such as ant colony optimization (ACO) and particle swarm optimization (PSO) are computationally intensive. Ant Brood Clustering (ABC) algorithm is one of the techniques in the family of ant algorithms. It is based on how ants cluster their brood or corpse into different piles. ABC has been efficiently used in solving the clustering problem in data analytics. However, they are computationally intensive. In this paper, we explore and evaluate eight parallel strategies of the ant brood clustering algorithm on Intel Xeon multi-core shared memory machine exploiting algorithm level and program level parallelism. A speedup of 25.3x is achieved with 56 hardware threads on dual 14 core Intel Xeon shared memory machine with coarse-grained, data level protection strategy. We show that contrary to other swarm intelligence techniques, fine grained parallelism is not suitable for this algorithm.
Date of Conference: 18-21 November 2018
Date Added to IEEE Xplore: 31 January 2019
ISBN Information:
Conference Location: Bangalore, India

References

References is not available for this document.