The main aim is to show that the actual approach for algorithms
complexity analysis based on Turing machine and its properties can be extended
trough the systems approach. It means that the actual existing paradigm, which
assumes that the analysis of algorithm behavior is done when the independence
between
input set and algorithm is assumed can be expanded by the approach that allows
a new view on possible phenomena that can occur during algorithmic processing.
Presented considerations allow to establish a connection between Turing
machines and physics of processing as a part of paradigm that considers
computer systems as the complex ones.