Modern computation problems arise that cannot be solved by increasing the number and quality of computers alone. In this work, we develop distributed computing methods. By applying knowledge from different areas of science, such as queueing theory and theory of computation, we create our own distributed computing system. The main idea behind our approach is the equality of hierarchy and decentralization, and generalization of calculations, covering more high-performance computation areas. Usage of peer-to-peer architecture requires a thoughtful design but offers several advantages, which are discussed in this work. To compose our system, we set a mathematical model of the target situation, draw attention to its weaknesses, and develop software principles to resolve them. After solving several design issues, we create simulation models to analyze the effectiveness of the future system. After analyzing the models, we select the next steps to improve and finalize our system. © 2020, Springer Nature Switzerland AG.