This paper presents an optimization of MPI communication, called Adaptive-CoMPI, based on runtime compression of MPI messages exchanged by applications. The technique developed can be used for any application, because its implementation is transparent for the user, and integrates different compression algorithms for both MPI collective and point-to-point primitives. Furthermore, compression is turned on and off and the most appropriate compression algorithms are selected at runtime, depending on the characteristics of each message, the network behavior, and compression algorithm behavior, following a runtime adaptive strategy. Our system can be optimized for a specific application, through a guided strategy, to reduce the runtime strategy overhead. Adaptive-CoMPI has been validated using several MPI benchmarks and real HPC applications. Results show that, in most cases, by using adaptive compression, communication time is reduced, enhancing application performance and scalability.
Balkanski, D., Trams, M. and Rehm, W. ( 2003). Heterogeneous Computing With MPICH/Madeleine and PACX MPI: A Critical Comparison.
2.
Bell, T., Powell, M. and J.H.R.A. (2009). The Canterbury Corpushttp://corpus.canterbury.ac.nz/index.html
3.
Burtscher, M. and Ratanaworabhan, P. ( 2009). FPC: A high-speed compressor for double-precision floating-point data. IEEE Transactions on Computers58(1): 18-31.
4.
Carretero, J., No, J., Park, S.S., Choudhary, A. and Chen, P. ( 1998). COMPASSION: A parallel I/O runtime system including chunking and compression for irregular applications. In Proceedings of the International Conference on High-Performance Computing and Networking, April, pp. 668-677.
5.
Carter, R., Ciotti, B., Fineberg, S. and Nitzberg, B. ( 1992). NHT-1I I/O benchmarks. Technical ReportRND-92-016. NAS Systems Division, NASA Ames.
6.
Coco, S., D’Arrigo, V. and D.G. (2000). A Rice-based lossless data compression system for space. In Proceedings of the 2000 IEEE Nordic Signal Processing Symposium, pp. 133-142.
7.
Davis, G., Lau, L., Young, R., Duncalfe, F. and Brebber, L. ( 1998). Parallel run-length encoding (RLE) compression-reducing I/O in dynamic environmental simulations. Int. J. of High Performance Computing Applications12(4): 396- 410 (Special Issue on I/O in Parallel Applications).
8.
Filgueira, R., Singh, D. E., A. C., Carretero, J. ( 2009). CoMPI: Enhancing MPI-based applications performance and scalability using compression. European PVM/ MPI.
9.
Gabriel, E., Fagg, G.E., Bosilca, G., Angskun, T., Dongarra, J.J., Squyres, J.M., Sahay, V., Kambadur, P., Barrett, B., Lumsdaine, A., Castain, R.H., Daniel, D.J., Graham, R.L. and Woodall, T.S. ( 2004). Open MPI: Goals, concept, and design of a next generation MPI implementation. In Proceedings, 11th European PVM/MPI Users’ Group Meeting, Budapest, Hungary, September, pp. 97-104.
10.
Garcia-Carballeira, F. and Alejandro Calderon, J.C. ( 1999). MiMPI: A multithread-safe implementation of MPI. In Recent Advances in Parallel Virtual Machine and Message Passing Interface, 6th European PVM/MPI Users’ Group Meeting, Springer, pp. 207-214.
11.
Gropp, W. and Lusk, E. ( 1997). Sowing MPICH: A case study in the dissemination of a portable environment for parallel scientific computing. Int. J. of Supercomputer Applications and High Performance Computing11(2): 103-114.
12.
Gropp, W. and Lusk, E.L. ( 1999). Reproducible measurements of MPI performance characteristics . In Proceedings of the 6th European PVM/MPI Users’ Group Meeting on Recent Advances in Parallel Virtual Machine and Message Passing Interface, London, UK, Springer-Verlag , pp. 11-18.
13.
Hockney, R.W. ( 1994). The communication challenge for MPP: Intel Paragon and Meiko CS-2. Parallel Computing20(3): 389-398.
14.
Ke, J., Burtscher, M. and Speight, E. ( 2004). Runtime compression of MPI messages to improve the performance and scalability of parallel applications. In SC ‘04, Proceedings of the 2004 ACM/IEEE conference on Supercomputing, 59.
15.
Keller, R. and M.L. (2005). Using PACX-MPI in metacomputing applications. In 18th Symposium Simulations technique, Erlangen, September, pp. 12-15.
16.
Knuth, D.E. ( 1985). Dynamic Huffman coding. J. Algorithms6(2): 163-180.
17.
Kumar, V.S., Nanjundiah, R., Thazhuthaveetil, M.J. and Govindarajan, R. ( 2008). Impact of message compression on the scalability of an atmospheric modeling application on clusters. Parallel Comput . 34(1): 1-16.
18.
Liu, J., Mamidala, A.R. and Panda, D.K. ( 2003). Fast and scalable MPI-level broadcast using InfiniBand’s hardware multicast support. In Proceedings of IPDPS.
19.
Loureiro, A., Gonzlez, J. and Pena, T.F. ( 2003). A parallel 3D semiconductor device simulator for gradual heterojunction bipolar transistors. Int. J. of Numerical Modelling: Electronic Networks, Devices and Fields16: 53-66.
20.
Majumdar, A. ( 1996). On evaluating and optimizing weights for weighted random pattern testing. IEEE Trans. Comput. 45(8): 904-916.
21.
Martin, R.P., Vahdat, A.M., Culler, D.E. and Anderson, T.E. ( 1997). Effects of communication latency, overhead, and bandwidth in a cluster architecture. SIGARCH Comput. Archit. News25(2): 85-97.
22.
Martinasso, M. and Méhaut, J.F. ( 2005). Prediction of communication latency over complex network behaviors on SMP clusters. In EPEW/WS-FM, pp. 172-186.
23.
Mourino, J., Martin, M., Doallo, R., Singh, D., Rivera, F. and Bruguera, J. ( 2004). The STEM-II air quality model on a distributed memory system.
24.
MPI Forum (1994). MPI: A Message-Passing Interface standard. Int. J. of Supercomputer Applications and High Performance Computing8(3/4): 165-414.
25.
Mucci, P.J., London, K. and Mucci, P.J. ( 1998). The MPBench Report. Technical report .
26.
Oberhumer, M.F.X.J. ( 2005). LZO real-time data compression library.
27.
Pichel, J.C., Singh, D.E. and Rivera, F.F. ( 2006). Image segmentation based on merging of sub-optimal segmentations . Pattern Recogn. Lett. 27(10): 1105-1116.
28.
Rashti, M.J. and Afsahi, A. ( 2008). Improving communication progress and overlap in MPI rendezvous protocol over RDMA-enabled interconnects. In HPCS ‘08, Proceedings of the 2008 22nd International Symposium on High Performance Computing Systems and Applications, Washington, DC, USA , IEEE Computer Society, pp. 95-101.
29.
Ratanaworabhan, P., Ke, J. and Burtscher, M. ( 2006). Fast lossless compression of scientific floating-point data. In DCC ‘06, Proceedings of the Data Compression Conference, Washington, DC, USA, IEEE Computer Society, pp. 133-142.
30.
Reussner, R., Sanders, P. and Träff, J.L. ( 2002). SKaMPI: a comprehensive benchmark for public benchmarking of MPI. Sci. Program. 10(1): 55-65.
31.
Thakur, R. ( 2006). Issues in developing a thread-safe MPI implementation . In Recent Advances in Parallel Virtual Machine and Message Passing Interface, 13th European PVM/MPI Users’ Group Meeting , Springer, pp. 12-21.
32.
Wunderlich, H.J. ( 1990). Multiple distributions for biased random test patterns . IEEE Trans. Computer-Aided Design of Integrated Circuits and Systems9(6): 584-593.
33.
Zigon, R. ( 1989). Run length encoding. j-DDJ14(2).
