Unordered Task-Parallel Augmented Merge Tree Construction

Contemporary scientific data sets require fast and scalable topological analysis to enable visualization, simplification and interaction. Within this field, parallel merge tree construction has seen abundant recent contributions, with a trend of decentralized, task-parallel or SMP-oriented algorithms dominating in terms of total runtime. However, none of these recent approaches computed complete merge trees on distributed systems, leaving this field to traditional divide & conquer approaches. This article introduces a scalable, parallel and distributed algorithm for merge tree construction outperforming the previously fastest distributed solution by a factor of around three. This is achieved by a task-parallel identification of individual merge tree arcs by growing regions around critical points in the data, without any need for ordered progression or global data structures, based on a novel insight introducing a sufficient local boundary for region growth.