ABSTRACTS OF ARTICLES OF THE JOURNAL "INFORMATION TECHNOLOGIES".
No. 12. Vol. 24. 2018

DOI: 10.17587/it.24.782-790

S. A. Inyutin, PhD, Professor, e-mail: inyutin_int@mail.ru, Moscow Aviation Institute (Nation Research University) (MAI)

The Aggregation of Number Systems for Multi-Digit Computational Processes

Designs and variants of combinations of external and internal number systems — mathematical bases of calculation methods in large computer ranges (multi-digit calculations) are introduced. Data formats for multi-digit calculations are systematized and substantiated. Studied design, containing internal and external positional power of the modular system. On the basis of code-independent Pepen and Lucas-Lehmer algorithms, code-dependent modular modifications of algorithms with less time complexity due to high memory costs are introduced. Linear spaces of modular vectors are investigated. The features and advantages of a modular system with a quadratic range and the corresponding parallel computer data formats for modular reconfigurable computing systems are revealed. The shown advantages justify the use of an external quadratic modular and internal positional number system within the machine range for the developed software tools for the organization of computational processes in large computer ranges on SIMD architecture computers.
Keywords: multiprocessor reconfigurable systems SIMD architecture, multi-row computational process, code-dependent algorithms, linear and quadratic complexity of algorithms, parallel computer formats of quadratic modular system

P. 782–790

To the contents