Parallel computing as a congestion game

TL;DR

This paper models parallel computing as a congestion game, enabling analysis of resource competition among processes to optimize performance and understand the impact of different optimality principles.

Contribution

It introduces a game-theoretical framework for analyzing parallel algorithms by representing resources as congestion game components.

Findings

Provides a new perspective on parallel algorithm analysis.

Enables comparative analysis of optimality principles.

Facilitates understanding of resource contention effects.

Abstract

Game-theoretical approach to the analysis of parallel algorithms is proposed. The approach is based on presentation of the parallel computing as a congestion game. In the game processes compete for resources such as core of a central processing unit and a communication subsystem. There are players, resources and payoffs (time delays) of players which depend on resources usage. Comparative analysis of various optimality principles in the proposed model may be performed.