Communication-Aware Supernode Shape

• Published in: IEEE transactions on parallel and distributed systems Vol. 20; no. 4; pp. 498 - 511
• Main Authors: Goumas, G., Drosinos, N., Koziris, N.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01-04-2009; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Bibliographies; Cartesian; Communication; Computational modeling; Computer architecture; Criteria; Data communications; Delay; Equivalence; I/O and Data Communications; Image processing; Load balancing and task assignment; Minimization; Minimization methods; Optimization; Parallel Architectures; Parallel processing; Parallel processors; Processor scheduling; Scheduling and task partitioning; Studies; Transformations; I/O and Data Communications; Data communications; Scheduling and task partitioning; Parallel Architectures; Load balancing and task assignment; Parallel processors
• ISSN: 1045-9219; 1558-2183
• DOI: 10.1109/TPDS.2008.114

In this paper we revisit the supernode-shape selection problem, that has been widely discussed in bibliography. In general, the selection of the supernode transformation greatly affects the parallel execution time of the transformed algorithm. Since the minimization of the overall parallel execution time via an appropriate supernode transformation is very difficult to accomplish, researchers have focused on scheduling-aware supernode transformations that maximize parallelism during the execution. In this paper we argue that the communication volume of the transformed algorithm is an important criterion, and its minimization should be given high priority. For this reason we define the metric of the per process communication volume and propose a method to minimize this metric by selecting a communication-aware supernode shape. Our approach is equivalent to defining a proper Cartesian process grid with MPI_Cart_Create, which means that it can be incorporated in applications in a straightforward manner. Our experimental results illustrate that by selecting the tile shape with the proposed method, the total parallel execution time is significantly reduced due to the minimization of the communication volume, despite the fact that a few more parallel execution steps are required.