Modeling of the Graf-dynamic machine architecture with computing cluster

Abstract

The model of the parallel architecture of problem-oriented processor has been developed. For the purpose of verification of delivered technical solutions the simulation of the architecture on the test cases with cluster GPU using CUDA technology has been provided. As a result, the model allowed to simulate of individual units and to explore the computational processes that occur in the studied architecture.

Keywords

• ассоциативный поиск
• семантическая сеть
• граф
• параллельные вычисления