Basic operations on BSN-Hypercube multiprocessors

被引:0
作者
Sun, Liting [1 ]
Tong, Chaonan [1 ]
机构
[1] Key Laboratory of Advanced Control of Iron and Steel Process (Ministry of Education), School of Automation Engineering, University of Science and Technology Beijing
来源
Journal of Computational Information Systems | 2014年 / 10卷 / 12期
关键词
Biswapped Network (BSN); Communication algorithms; Hypercube; Prefix sum; Sparse sorting; Topological property;
D O I
10.12733/jcis10664
中图分类号
学科分类号
摘要
Hypercube is the most famous, common, and effective interconnection network, hence it has become to be the preferred topology of the parallel processing and parallel computing system. On the other hand, Biswapped network (BSN) provides an efficient two-level structure for large scalable, inheritable, modular architecture. In this paper, we present BSN-Hypercube as a two-tier architecture that takes the advantages of both BSN and Hypercube. Topological property and many basic operations are investigated. The algorithms include unicast, multicast, data sum, prefix sum, and sparse sorting. In contrast with other two-tier architectures, the results show that the algorithms that run on BSN-Hypercube are much faster, and BSN-Hypercube is more competitive. 1553-9105/Copyright © 2014 Binary Information Press.
引用
收藏
页码:5211 / 5218
页数:7
相关论文
共 10 条
[1]  
Xiao W., Biswapped networks and their topological properties, SNPD 2007: 8th ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing, (2007)
[2]  
Chen W., Xiao W., Topological properties of Biswapped Networks (BSNs): Node symmetry and maximal fault tolerance, Jisuanji Xuebao/Chinese Journal of Computers, 33, 5, pp. 822-832, (2010)
[3]  
Wei W., Xiao W., Fault tolerance in the biswapped network, 8th International Conference on Algorithms and Architectures for Parallel Processing, (2008)
[4]  
Ye H., Xiao W., Wu J., Broadcasting on the BSN-hypercube network, 2009 2nd International Conference on Information and Computing Science, ICIC 2009, (2009)
[5]  
Ye H., Xiao W., Zhu X., Some parallel algorithms and broadcasting on the BSN-hypercube network, Journal of Information and Computational Science, 6, 2, pp. 629-635, (2009)
[6]  
Wei W., Xiao W., Algorithms of basic communication operation on the biswapped network, 8th International Conference on Computational Science, ICCS 2008, (2008)
[7]  
Wei W., Xiao W., Matrix multiplication on the biswapped-mesh network, SNPD 2007: 8th ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing, (2007)
[8]  
Sun L., Tong C., Su H., New mapping scheme of matrix algorithms on the Biswapped network, Journal of Computational Information Systems, 9, 13, pp. 5371-5378, (2013)
[9]  
Jana P.K., Multi-mesh of trees with its parallel algorithms, Journal of Systems Architecture, 50, 4, pp. 193-206, (2004)
[10]  
Islam R., Afroz N., Bandyopadhyay S., Sinha B.P., Computational geometry on optical multitrees (OMULT) computer system, CCCG, pp. 150-154, (2005)