Clustering Graphs for GPU Processing

Abstract

This article discusses the multilevel clustering of graphs used for data processing in the GPU. The quality of graph partitioning can have a significant impact on the overall performance of the software. Therefore, it is very important to quickly find the correct partition of the graph into subgraphs. The article describes the application areas of clustering, features and types of clustering, and also draws conclusions about the relevance of using graph clustering in modern spheres of human activity.