Abstract
Graph clustering is pervasive in emerging “big data” applications, and is known to be quite challenging to implement on distributed memory systems. In this work, we design and implement scalable distributed-memory algorithms for peer pressure clustering using the sparse matrix infrastructures of Combinatorial BLAS, where the peer pressure clustering algorithm is represent as sparse matrix computations. For settling ties, which is the most time-consuming step in this algorithm, we design a matrix-based algorithm and provide two parallel implementations. One is based on MPI model, and the other is a hybrid programming with MPI and OpenMP. Relying on matrix algebra building blocks, our algorithm exposes a high degree of parallelism and good scalability on distributed-memory platforms. For a real instance, when the input is a permuted R-MAT graph of scale 21 with self-loops added, our MPI implementation achieves up to 809.6x speedup on 1024 cores of a Dawning supercomputer, and the hybrid implementation with MPI and OpenMP obtains 949.5x speedup on 2048 cores of the same computer.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Shah, V.B.: An interactive system for combinatorial scientific computing with an emphasis on programmer productivity, Dissertation. University of California, Santa Barbara (2007)
Techentin, R.W., et al.: Implementing iterative algorithms with SPARQL. Edbt/icdt workshops 2014, pp. 216–223 (2014)
Deweese, K., et al.: Graph clustering in SPARQL. In: SIAM Workshop on Network Science, vol. 34, pp. 930–941 (2013)
Gilbert, J.R., Reinhardt, S., Shah, V.B.: High-performance graph algorithms from parallel sparse matrices. In: Kågström, B., Elmroth, E., Dongarra, J., Waśniewski, J. (eds.) PARA 2006. LNCS, vol. 4699, pp. 260–269. Springer, Heidelberg (2007). doi:10.1007/978-3-540-75755-9_32
Harary, F.: Graph Theory. Addison-Wesley, Boston (1969)
Buluç, A., Gilbert, J.R.: Highly parallel sparse matrix-matrix multiplication. Computer Science, abs/1006.2183 (2010)
Buluç, A., Gilbert, J.R.: Parallel sparse matrix-matrix multiplication and indexing: Implementation and experiments. SIAM J. Sci. Comput. 34(4), C170–C191 (2012)
Buluç, A., Gilbert, J.R.: Challenges and advances in parallel sparse matrix-matrix multiplication. In: 37th International Conference on Parallel Processing, ICPP 2008, pp. 503–510. IEEE (2008)
Buluç, A., Gilbert, J.R.: On the representation and multiplication of hypersparse matrices. In: IEEE International Symposium on Parallel and Distributed Processing, IPDPS 2008, pp. 1–11. IEEE (2008)
Buluç, A., Gilbert, J.R.: The combinatorial BLAS: design, implementation, and applications. Int. J. High Perform. Comput. Appl. 25(4), 496–509 (2011)
Chakrabarti, D., Zhan, Y., Faloutsos, C.: R-MAT: a recursive model for graph mining. In: SDM, vol. 4, pp. 442–446 (2004)
Jeremy, K., Gilbert, J. (eds.): Graph Algorithms in the Language of Linear Algebra, vol. 22. SIAM, Philadelphia (2011)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Chen, J., Zou, P. (2017). Parallel Peer Pressure Clustering Algorithm Based on Linear Algebra Computation. In: Dou, Y., Lin, H., Sun, G., Wu, J., Heras, D., Bougé, L. (eds) Advanced Parallel Processing Technologies. APPT 2017. Lecture Notes in Computer Science(), vol 10561. Springer, Cham. https://doi.org/10.1007/978-3-319-67952-5_10
Download citation
DOI: https://doi.org/10.1007/978-3-319-67952-5_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-67951-8
Online ISBN: 978-3-319-67952-5
eBook Packages: Computer ScienceComputer Science (R0)
