Affiliation:
1. University of Southampton, UK
2. Electronics and Computer Science, University of Southampton, UK
Abstract
A multi-core cluster is a cluster composed of numbers of nodes where each node has a number of processors, each with more than one core within each single chip. Cluster nodes are connected via an interconnection network. Multi-cored processors are able to achieve higher performance without driving up power consumption and heat, which is the main concern in a single-core processor. A general problem in the network arises from the fact that multiple messages can be in transit at the same time on the same network links. This chapter considers the communication latencies of a multi core multi cluster architecture, investigated using simulation experiments and measurements under various working conditions.
Reference71 articles.
1. Abawajy, J. H., & Dandamudi, S. P. (2003). Parallel job scheduling on multicluster computing system. In Proceedings of the 2003 IEEE International Conference on Cluster Computing.
2. On the Performance of MPI-OpenMP on a 12 Nodes Multi-core Cluster
3. Admin. (2009). TOP500 Highlights - November 2009. Retrieved from http://www.top500.org/lists/2009/11/highlights
4. Admin. (2010). What is cluster computing. Retrieved from http://www.ccgrid.org/tag/cluster-computing
5. Admin. (2013). Top500 Supercomputer Sites.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Hardware Interrupt and CPU Contention aware CPU/GPU Co-Scheduling on Multi-Cluster System;2022 5th International Conference on Information and Computer Technologies (ICICT);2022-03