Abstract
In parallel programming when a graph is distributed among processors it is useful to introduce consecutive local numeration of vertices for fast access to vertices and their adjacent vertices. The problem of mapping of global adjacency lists to local numeration arises. A graph can be distributed among processors after predecomposition or some vertices can be moved to other processors during partitioning. So vertex global numbers on a processor can be any and in any order. The algorithm devised in the parallel partitioning tool GridSpiderPar works with any local numeration, even with nonoptimal one. It can be applied to a whole graph or to a part of the graph. This algorithm allowed to reduce execution time of mapping of global adjacency lists to local numeration in a distributed graph in 18 times and the time of adding vertices to the graph while redistributing groups of bad subdomains – in 32 times on a tetrahedral mesh with 2·108 vertices. All these algorithms are parts of the parallel incremental algorithm for graph partitioning from the GridSpiderPar package. The proposed algorithm can also be used when some information about vertices is received from the other processors. It helps to find these vertices in the local graph faster. In common it replaces a sequence of searches in a large array with one looking through two sorted arrays for coincidences and can be applied in such algorithms.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Golovchenko, E., Dorofeeva, E., Gasilova, I., Boldarev, A.: Numerical experiments with new algorithms for parallel decomposition of large computational meshes. Parallel Computing: Accelerating Computational Science and Engineering (CSE). Advances in Parallel Computing, vol. 25, pp. 441–450. IOS Press (2014)
Golovchenko, E.N., Kornilina, M.A., Yakobovskiy, M.V.: Algorithms in the parallel partitioning tool GridSpiderPar for large mesh decomposition. In: Proceedings of the 3rd International Conference on Exascale Applications and Software (EASC 2015), pp. 120–125. University of Edinburgh (2015)
Karypis, G., Schloegel, K.: Parmetis. Parallel Graph Partitioning and Sparse Matrix Ordering Library. Version 4.0. University of Minnesota, Department of Computer Science and Engineering. Minneapolis, MN 55455, (2013). karypis@cs.umn.edu
Mikhaylov, S.V.: Principles of creating code for solving the aerodynamics and aeroacoustics. Math. Models Comput. Simul. 29(9), 49–61 (2017)
Devine, K.D., Boman, E.G., Heaphy, R.T., Bisseling, R.H., Catalyurek, U.V.: Parallel hypergraph partitioning for scientific computing. In: Proceedings 20th IEEE International Parallel & Distributed Processing Symposium, p. 10 (2010)
Walshaw, C., Cross, M., Everett, M.G.: Parallel dynamic graph partitioning for adaptive unstructured meshes. J. Parallel Distrib. Comput. 47, 102–108 (1997)
Zhang, Y., Li, D., Zhang, C., Wang, J., Liu, L.: GraphA: efficient partitioning and storage for distributed graph computation. Trans. Serv. Comput. 14(8), 155–166 (2016)
Leis, V., Kemper, A., Neumann, T.: The adaptive radix tree: artful indexing for main-memory databases. In: European Conference on Computer Systems, pp. 38–49 (2013)
Soukov, S.A., Gorobets, A.V., Bogdanov, P.B.: Portable solution for modeling of compressible turbulent flows on whatever hybrid systems. Math. Models Comput. Simul. 10(2), 135–144 (2018)
Golovchenko, E.N.: Computational mesh partitioning in numerical solution of continuum mechanics problems on high-performance computing systems. – Moscow: Keldysh Institute of Applied Mathematics RAS, Ph. D. thesis, Candidate of Science in Physics and Mathematics (2014)
Donald, K.: The Art of Computer Programming, v.3. Sorting and Searching 3rd ed., Addison Wesley, Boston, MA (1997)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Golovchenko, E. (2023). An Algorithm for Mapping of Global Adjacency Lists to Local Numeration in a Distributed Graph in the GridSpiderPar Tool. In: Voevodin, V., Sobolev, S., Yakobovskiy, M., Shagaliev, R. (eds) Supercomputing. RuSCDays 2023. Lecture Notes in Computer Science, vol 14389. Springer, Cham. https://doi.org/10.1007/978-3-031-49435-2_10
Download citation
DOI: https://doi.org/10.1007/978-3-031-49435-2_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-49434-5
Online ISBN: 978-3-031-49435-2
eBook Packages: Computer ScienceComputer Science (R0)