Understanding and Modeling Lossy Compression Schemes on HPC Scientific Data

Lu, Tao; Liu, Qing Gary; He, Xubin; Luo, Huizhang; Suchyta, Eric; Choi, Jong Youl; Podhorszki, Norbert; Klasky, Scott; Wolf, Matthew; Liu, Tong; Qiao, Zhenbo

doi:10.1109/IPDPS.2018.00044

Understanding and Modeling Lossy Compression Schemes on HPC Scientific Data

Conference · 01 May 2018

DOI:https://doi.org/10.1109/IPDPS.2018.00044· OSTI ID:1468061

Lu, Tao ^[1]; Liu, Qing Gary ^[2]; He, Xubin ^[3]; Luo, Huizhang ^[1]; ^[2]; ^[2]; ^[2]; ^[2]; ^[2]; Liu, Tong ^[3]; Qiao, Zhenbo ^[1]

New Jersey Institute of Technology
ORNL
Temple University

Scientific simulations generate large amounts of floating-point data, which are often not very compressible using the traditional reduction schemes, such as deduplication or lossless compression. The emergence of lossy floating-point compression holds promise to satisfy the data reduction demand from HPC applications; however, lossy compression has not been widely adopted in science production. We believe a fundamental reason is that there is a lack of understanding of the benefits, pitfalls, and performance of lossy compression on scientific data. In this paper, we conduct a comprehensive study on state-of-the-art lossy compression, including ZFP, SZ, and ISABELA, using real and representative HPC datasets. Our evaluation reveals the complex interplay between compressor design, data features and compression performance. The impact of reduced accuracy on data analytics is also examined through a case study of fusion blob detection, offering domain scientists with the insights of what to expect from fidelity loss. Furthermore, the trial and error approach to understanding compression performance involves substantial compute and storage overhead. To this end, we propose a sampling based estimation method that extrapolates the reduction ratio from data samples, to guide domain scientists to make more informed data reduction decisions.

View Conference

Research Organization:: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)

Sponsoring Organization:: USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)

DOE Contract Number:: AC05-00OR22725

OSTI ID:: 1468061

Resource Relation:: Conference: 32nd IEEE International Parallel and Distributed Processing Symposium (IPDPS 2018) - Vancouver, , Canada - 5/21/2018 4:00:00 AM-5/25/2018 4:00:00 AM

Country of Publication:: United States

Language:: English

References (23)

A Mathematical Theory of Communication Shannon, C. E. Bell System Technical Journal, Vol. 27, Issue 3 https://doi.org/10.1002/j.1538-7305.1948.tb01338.x	journal	July 1948
Parallel Tensor Compression for Large-Scale Scientific Data Austin, Woody; Ballard, Grey; Kolda, Tamara G. 2016 IEEE International Parallel and Distributed Processing Symposium (IPDPS) https://doi.org/10.1109/IPDPS.2016.67	conference	May 2016
The predictability of data values Sazeides, Y.; Smith, J. E. Proceedings of 30th Annual International Symposium on Microarchitecture https://doi.org/10.1109/MICRO.1997.645815	conference	January 1997
Differential FCM: increasing value prediction accuracy by improving table usage efficiency Goeman, B.; Vandierendonck, H.; de Bosschere, K. HPCA-7 - 7th IEEE Symposium on High Performance Computer Architecture, Proceedings HPCA Seventh International Symposium on High-Performance Computer Architecture https://doi.org/10.1109/HPCA.2001.903264	conference	January 2001
Spline Functions in Data Analysis Wold, Svante Technometrics, Vol. 16, Issue 1 https://doi.org/10.2307/1267485	journal	February 1974
FPC: A High-Speed Compressor for Double-Precision Floating-Point Data Burtscher, Martin; Ratanaworabhan, Paruj IEEE Transactions on Computers, Vol. 58, Issue 1 https://doi.org/10.1109/TC.2008.131	journal	January 2009
Embedded image coding using zerotrees of wavelet coefficients Shapiro, J. M. IEEE Transactions on Signal Processing, Vol. 41, Issue 12 https://doi.org/10.1109/78.258085	journal	January 1993
Significantly Improving Lossy Compression for Scientific Data Sets Based on Multidimensional Prediction and Error-Controlled Quantization Tao, Dingwen; Di, Sheng; Chen, Zizhong 2017 IEEE International Parallel and Distributed Processing Symposium (IPDPS) https://doi.org/10.1109/IPDPS.2017.115	conference	May 2017
Hello ADIOS: the challenges and lessons of developing leadership class I/O frameworks: HELLO ADIOS Liu, Qing; Logan, Jeremy; Tian, Yuan Concurrency and Computation: Practice and Experience, Vol. 26, Issue 7 https://doi.org/10.1002/cpe.3125	journal	August 2013
Monotone B-Spline Smoothing He, Xuming; Shi, Peide Journal of the American Statistical Association, Vol. 93, Issue 442 https://doi.org/10.2307/2670115	journal	June 1998
On the Complexity of Finite Sequences Lempel, A.; Ziv, J. IEEE Transactions on Information Theory, Vol. 22, Issue 1 https://doi.org/10.1109/TIT.1976.1055501	journal	January 1976
A study on data deduplication in HPC storage systems Meister, Dirk; Kaiser, Jurgen; Brinkmann, Andre 2012 International Conference for High Performance Computing, Networking, Storage and Analysis https://doi.org/10.1109/SC.2012.14	conference	November 2012
Exploration of Lossy Compression for Application-Level Checkpoint/Restart Sasaki, Naoto; Sato, Kento; Endo, Toshio 2015 IEEE International Parallel and Distributed Processing Symposium (IPDPS) https://doi.org/10.1109/IPDPS.2015.67	conference	May 2015
Sur les fonctions convexes et les inégalités entre les valeurs moyennes Jensen, J. L. W. V. Acta Mathematica, Vol. 30, Issue 0 https://doi.org/10.1007/BF02418571	journal	January 1906
Fast Error-Bounded Lossy HPC Data Compression with SZ Di, Sheng; Cappello, Franck 2016 IEEE International Parallel and Distributed Processing Symposium (IPDPS) https://doi.org/10.1109/IPDPS.2016.11	conference	May 2016
Counting YouTube videos via random prefix sampling Zhou, Jia; Li, Yanhua; Adhikari, Vijay Kumar Proceedings of the 2011 ACM SIGCOMM conference on Internet measurement conference https://doi.org/10.1145/2068816.2068851	conference	November 2011
Fixed-Rate Compressed Floating-Point Arrays Lindstrom, Peter IEEE Transactions on Visualization and Computer Graphics, Vol. 20, Issue 12 https://doi.org/10.1109/TVCG.2014.2346458	journal	December 2014
The JPEG still picture compression standard Wallace, G. K. IEEE Transactions on Consumer Electronics, Vol. 38, Issue 1 https://doi.org/10.1109/30.125072	journal	January 1992
A Comprehensive Study of the Past, Present, and Future of Data Deduplication Xia, Wen; Jiang, Hong; Feng, Dan Proceedings of the IEEE, Vol. 104, Issue 9 https://doi.org/10.1109/JPROC.2016.2571298	journal	September 2016
Canopus: A Paradigm Shift Towards Elastic Extreme-Scale Data Analytics on HPC Storage Lu, Tao; Suchyta, Eric; Pugmire, Dave 2017 IEEE International Conference on Cluster Computing (CLUSTER) https://doi.org/10.1109/CLUSTER.2017.62	conference	September 2017
A universal algorithm for sequential data compression Ziv, J.; Lempel, A. IEEE Transactions on Information Theory, Vol. 23, Issue 3 https://doi.org/10.1109/TIT.1977.1055714	journal	May 1977
MaDaTS: Managing Data on Tiered Storage for Scientific Workflows Ghoshal, Devarshi; Ramakrishnan, Lavanya Proceedings of the 26th International Symposium on High-Performance Parallel and Distributed Computing - HPDC '17 https://doi.org/10.1145/3078597.3078611	conference	January 2017
ArrayUDF: User-Defined Scientific Data Analysis on Arrays Dong, Bin; Wu, Kesheng; Byna, Surendra Proceedings of the 26th International Symposium on High-Performance Parallel and Distributed Computing - HPDC '17 https://doi.org/10.1145/3078597.3078599	conference	January 2017

Similar Records

Accelerating Relative-error Bounded Lossy Compression for HPC datasets with Precomputation-Based Mechanisms

Conference · 2019 · OSTI ID:1515554

Zou, Xiangyu; Lu, Tao; Xia, Wen; +5 more

Performance Optimization for Relative-Error-Bounded Lossy Compression on Scientific Data

Journal Article · 2020 · IEEE Transactions on Parallel and Distributed Systems · OSTI ID:1603491

Zou, Xiangyu; Lu, Tao; Xia, Wen; +6 more

Estimating Lossy Compressibility of Scientific Data Using Deep Neural Networks

Journal Article · 2020 · IEEE Letters of the Computer Society · OSTI ID:1824994

Qin, Zhenlu; Wang, Jinzhen; Liu, Qing; +4 more

Understanding and Modeling Lossy Compression Schemes on HPC Scientific Data

Citation Formats

References (23)

Similar Records

Related Subjects