Distributed processing is a new concept considered as a solution to the cost problem relating to scaling parallel machines. Any problem given to a distributed system of n-processors can be better solved by a single n-processor parallel machine. However it is not always best to use all processors on a given machine in a distributed system. Thus, for example, it is only sometimes best to spread an n + 1 processor algorithm onto all n processors of an n-processor machine and have the other processor on any p-processor machine. We use the calculation of permutations as an example of a problem best divided up more carefully. Since many classes of operations on finite data sets can be mapped to subgroups of permutation groups, permutations hold a special place in any discussion on the manipulation, interaction and transmission of data. We briefly discuss the distribution of three different algorithms for calculating permutations.
Journal: TechConnect Briefs
Volume: 2, Technical Proceedings of the 2001 International Conference on Computational Nanoscience and Nanotechnology
Published: March 19, 2001
Pages: 200 - 203
Industry sector: Advanced Materials & Manufacturing
Topics: Informatics, Modeling & Simulation