Static prediction of parallel computation graphs

Static performance prediction of skeletal parallel programs

We demonstrate that the run time of implicitly parallel programs can be statically predicted with considerable accuracy when expressed within the constraints of a skeletal, shapely parallel programming language. Our work constitutes the first completely static system to account for both computation and communication in such a context. We present details of our language and its BSP implementatio...

Static and Dynamic Parallel Computation of Connected Components

The connected components of an undirected graph are determined in the EREW model with p processors, using a parallel extension of the “sparsification” technique. A previous algorithm due to Kruskal et al. is optimal forp < ml/(*+&) < n/2. We show that our algorithm is work-optimal for an extended range of values. Using the “sparsification” tree data structure we provide also a dynamic algorithm...

Optimal Replication of Series-Parallel Graphs for Computation-Intensive Applications

PowerGraph: Distributed Graph-Parallel Computation on Natural Graphs

Large-scale graph-structured computation is central to tasks ranging from targeted advertising to natural language processing and has led to the development of several graph-parallel abstractions including Pregel and GraphLab. However, the natural graphs commonly found in the real-world have highly skewed power-law degree distributions, which challenge the assumptions made by these abstractions...

Parallel computation on interval graphs: algorithms and experiments

This paper describes efficient coarse-grained parallel algorithms and implementations for a suite of interval graph problems. Included are algorithms requiring only a constant number of communication rounds for connected components, maximum weighted clique, and breadth-first-search and depth-first-search trees, as well as O(log p) communication rounds algorithms for optimization problems such a...