Дослідження використання GRID-систем для масового аналізу шахових партій

Abstract

This paper investigates the application of GRID systems for large-scale analysis of chess games using modern approaches to the organization of computational processes. Modern chess information resources contain millions of games, the analysis of which is important for training chess players, forming opening databases, statistical research, and training intelligent systems. The use of sequential algorithms for such processing leads to significant time costs and does not allow for effective scaling of calculations. The proposed approach is based on the disruption of analysis tasks among the nodes of a GRID system, where each node performs independent analysis of individual games or positions using standard chess engines. The paper examines the principles of organizing such analysis, discusses the architectural features of GRID systems, and justifies their suitability for chess analytics involving large volumes of data.