GPSS Modelling of Memristor-Based Computing Systems

When developing artificial intelligence systems based on memristors, it is necessary to determine which values of the main indicators of the quality of their functioning, such as accuracy, fault tolerance, reliability, performance, speed, power consumption, etc. will eventually be achieved. These parameters can be estimated using simulation modeling at different levels of the structural and functional hierarchy of the system. This paper shows that GPSS is a suitable tool for solving this problem at the system level. During the experiments in the GPSS language, models of distributed computing systems based on memristors with one and several data sources and with one and several data processing units were developed.