Abstract
Communication performance plays a crucial role in both the scalability and the time-to-solution of parallel applications. The share of links in modern high-performance computer networks inevitably introduces contention for communications involving multiple point-to-point messages, thus hinders their performance. Passive contention reduction such as the congestion control of the networks can mitigate network contention but with extra protocol cost, while application-level active contention reduction such as topology mapping techniques can only reduce contention of applications with static communication patterns. In this paper, we explore a different approach to actively reduce network contention through a congestion-avoiding message scheduling algorithm, namely CAMS. CAMS determines how to inject the messages in groups to reduce contention just in time before injecting them into the network, thus it is useful in applications with dynamic communication patterns. Experiments with a 2D halo-exchange benchmark on the Tianhe-2A supercomputer shows that it can improve communication performance up to 27% when messages get large. The proposed approach can be used in conjunction with topology mapping to further improve communication performance.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Agarwal, T., Sharma, A., Laxmikant, A., Kalé, L.V.: Topology-aware task mapping for reducing communication contention on large parallel machines. In: Proceedings 20th IEEE International Parallel and Distributed Processing Symposium, pp. 10-pp. IEEE (2006)
Alverson, R., Roweth, D., Kaplan, L.: The Gemini system interconnect. In: 2010 18th IEEE Symposium on High Performance Interconnects, pp. 83–87. IEEE (2010)
InniBand Trade Association: Inniband architecture specification, vol. 1 & 2, release 1.2, October 2004
Bhatelé, A., Bohm, E., Kalé, L.V.: Optimizing communication for Charm++ applications by reducing network contention. Concurr. Comput. Pract. Exp. 23(2), 211–222 (2011)
Chen, D., et al.: The IBM Blue Gene/Q interconnection fabric. IEEE Micro 32(1), 32–43 (2011)
Doi, J., Negishi, Y.: Overlapping methods of all-to-all communication and FFT algorithms for torus-connected massively parallel supercomputers. In: Proceedings of the 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2010, pp. 1–9. IEEE (2010)
Escudero-Sahuquillo, J., et al.: A new proposal to deal with congestion in infiniband-based fat-trees. J. Parallel Distrib. Comput. 74(1), 1802–1819 (2014)
Faanes, G., et al.: Cray cascade: a scalable HPC system based on a Dragonfly network. In: Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis, p. 103. IEEE Computer Society Press (2012)
Gomez, C., Gilabert, F., Gomez, M.E., López, P., Duato, J.: Deterministic versus adaptive routing in fat-trees. In: 2007 IEEE International Parallel and Distributed Processing Symposium, pp. 1–8. IEEE (2007)
Gran, E.G., et al.: First experiences with congestion control in infiniband hardware. In: 2010 IEEE International Symposium on Parallel and Distributed Processing (IPDPS), pp. 1–12. IEEE (2010)
Gustafson, J.L.: Reevaluating Amdahl’s law. Commun. ACM 31(5), 532–533 (1988)
Hoefler, T., Jeannot, E., Mercier, G.: An overview of topology mapping algorithms and techniques in high-performance computing, chap. 5, pp. 73–94. Wiley, Hoboken (2014). https://doi.org/10.1002/9781118711897.ch5
Hoefler, T., Snir, M.: Generic topology mapping strategies for large-scale parallel architectures. In: Proceedings of the International Conference on Supercomputing, pp. 75–84. ACM (2011)
Jain, N., Bhatele, A., Robson, M.P., Gamblin, T., Kale, L.V.: Predicting application performance using supervised learning on communication features. In: Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis, p. 95. ACM (2013)
Jiang, N., Becker, D.U., Michelogiannakis, G., Dally, W.J.: Network congestion avoidance through speculative reservation. In: IEEE International Symposium on High-Performance Computer Architecture, pp. 1–12. IEEE (2012)
Jiang, N., Dennison, L., Dally, W.J.: Network endpoint congestion control for fine-grained communication. In: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015, pp. 1–12. IEEE (2015)
Kamil, S., Oliker, L., Pinar, A., Shalf, J.: Communication requirements and interconnect optimization for high-end scientific applications. IEEE Trans. Parallel Distrib. Syst. 21(2), 188–202 (2009)
Lavrijsen, W., Iancu, C.: Application level reordering of remote direct memory access operations. In: 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS), pp. 988–997. IEEE (2017)
Lavrijsen, W., Iancu, C., Pan, X.: Improving network throughput with global communication reordering. In: 2018 IEEE International Parallel and Distributed Processing Symposium (IPDPS), pp. 266–275. IEEE (2018)
Luo, M., Panda, D.K., Ibrahim, K.Z., Iancu, C.: Congestion avoidance on manycore high performance computing systems. In: Proceedings of the 26th ACM International Conference on Supercomputing, pp. 121–132. ACM (2012)
Madduri, K., et al.: Gyrokinetic toroidal simulations on leading multi- and manycore HPC systems. In: Proceedings of 2011 International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2011, pp. 1–12. IEEE (2011)
Márquez, C., César, E., Sorribes, J.: A load balancing schema for agent-based SPMD applications. In: Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications (PDPTA), p. 12. The Steering Committee of the World Congress in Computer Science, Computer Engineering and Applied Computing (WorldComp) (2013)
Oliker, L., et al.: Scientific application performance on candidate petascale platforms. In: 2007 IEEE International Parallel and Distributed Processing Symposium, pp. 1–12. IEEE (2007)
Pfister, G., et al.: Solving hot spot contention using infiniband architecture congestion control. In: Proceedings HP-IPC 2005, p. 6 (2005)
Valiant, L.: A bridging model for parallel computation. Commun. ACM 33(8) (1990). https://doi.org/10.1145/79173.79181
Vetter, J.S., Mueller, F.: Communication characteristics of large-scale scientific applications for contemporary cluster architectures. In: Proceedings 16th International Parallel and Distributed Processing Symposium, pp. 10-pp. IEEE (2001)
Zahavi, E.: D-Mod-K routing providing non-blocking traffic for shift permutations on real life fat trees. CCIT Report 776, 840 (2010)
Zahavi, E., Johnson, G., Kerbyson, D.J., Lang, M.: Optimized infiniband TM fat-tree routing for shift all-to-all communication patterns. Concurr. Comput. Pract. Exp. 22(2), 217–231 (2010)
Acknowledgement
The authors would like to thank the National Supercomputer Center in Guangzhou for providing the experimental platform and tremendous help on usage of the Tianhe-2A supercomputer. This research was supported partially by Science Challenge Project (No. TZ2016002), National Key R&D Program (No. 2016YFB0201300) and Defense Industrial Technology Development Program (C1520110002). The authors also thank the reviewers for their helpful comments.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Peng, J., Yang, Z., Liu, Q. (2020). Improving Performance of Batch Point-to-Point Communications by Active Contention Reduction Through Congestion-Avoiding Message Scheduling. In: Wen, S., Zomaya, A., Yang, L. (eds) Algorithms and Architectures for Parallel Processing. ICA3PP 2019. Lecture Notes in Computer Science(), vol 11944. Springer, Cham. https://doi.org/10.1007/978-3-030-38991-8_27
Download citation
DOI: https://doi.org/10.1007/978-3-030-38991-8_27
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-38990-1
Online ISBN: 978-3-030-38991-8
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)