Data decomposition for formation aggregation values of hypercube in multiprocessor parallel computing systems
R Uskenbayeva, N Mukazhanov
COMPUTER MODELLING & NEW TECHNOLOGIES 2015 19(3A) 17-23
Department of Computer Science and Software Engineering, International University of Information Technologies 34 «А»/8 «А» Manas Str./Zhandosov Str., Almaty, Kazakhstan
In this paper possibilities of aggregational values calculation is considered. Aggregational values are the main element of multidimensional operative analytical processing. The main reason of using parallel computing systems in data processing is to increase productivity level. Although, parallel computing systems cannot be used in processing all data types. Data processing algorithms and processing data should be gradually adapted to parallel computing systems’ usage. In this regard, data decomposition for formation aggregational values in parallel computing systems in data operative analyzing is considered in this paper. In order to identify dependence between data during the process of decomposition Bernstein's conditions are used. At the same time implemented course calculation of from 1-dimension to n-dimension and parallel computation of course interactions will also be considered.