A distributed-memory algorithm for connected components labeling of simulation data

Abstract

This chapter describes a data-parallel, distributed-memory algorithm for identifying and labeling the connected sub-meshes within a three-dimensional mesh. The identification task is challenging in a distributed-memory setting because connectivity is transitive and the cells composing a sub-mesh may span many processors. The algorithm employs a multi-stage application of the Union-find algorithm and a spatial partitioning scheme to efficiently merge information across processors and to produce a global labeling of connected sub-meshes. Marking each vertex with its corresponding sub-mesh label allows mesh features to be isolated based on topology, enabling important analysis capabilities. The algorithm performs well in parallel; results are presented from a weak scaling study with concurrency levels up to 2,197 cores and meshes containing over two billion cells. This chapter is an extension of previous work by Harrison et al. (Data-parallel mesh connected components labeling and analysis. In: EuroGraphics Symposium on Parallel Graphics and Visualization (EGPGV), pp. 131–140, April 2011). It contains significant algorithmic improvements over the previous version, improved exploration of key bottlenecks in the algorithm, and improved clarity of presentation.