Compressão de Dados para Redução de E/S em Simulações Sísmicas de Alto Desempenho
Resumo
Simulações sísmicas de alta fidelidade geram grandes volumes de dados, tornando a E/S um gargalo crítico em aplicações de HPC. Este trabalho avalia o impacto da compressão de dados na redução desse gargalo, aplicando ferramentas de compressão sobre conjuntos de dados gerados por um modelo sísmico em diferentes escalas. A análise considera taxa de compressão, tempo de compressão e descompressão, e impacto no tempo total de execução. Os resultados mostram que o fpzip oferece maior compressão, enquanto o lz4 apresenta melhor desempenho em tempo de execução, e o zfp fornece um equilíbrio entre ambos. Também discutimos os desafios para aceleração em GPU, destacando a maturidade da biblioteca nvCOMP para GPUs NVIDIA e a ausência de soluções equivalentes para o ecossistema AMD, indicando caminhos para futuras otimizações em ambientes heterogêneos.
Referências
AMD (2025). Amd rocm documentation. [link]. Accessed: 2025-06-20.
Azami, N., Fallin, A., and Burtscher, M. (2025). Efficient lossless compression of scientific floating-point data on cpus and gpus. In Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 1, pages 395–409.
Barbosa, C. H. and Coutinho, A. L. (2022). Multi-gpu 3-d reverse time migration with minimum i/o. In Latin American High Performance Computing Conference, pages 160–173. Springer.
Chen, X., Tian, J., Beaver, I., Freeman, C., Yan, Y., Wang, J., and Tao, D. (2023a). Fcbench: Cross-domain benchmarking of lossless compression for floating-point data. Proceedings of the VLDB Endowment, 17(6):1418–1431.
Chen, X., Tian, J., Beaver, I., Freeman, C., Yan, Y., Wang, J., and Tao, D. (2024). Fcbench: Cross-domain benchmarking of lossless compression for floating-point data. Proceedings of the VLDB Endowment, 17(6):1418–1431.
Chen, Z., Di, S., Tao, D., and Cappello, F. (2023b). Beyond compression: A comprehensive evaluation of lossless floating-point compression. Journal of Parallel and Distributed Computing.
Chen, Z., Tao, D., and Cappello, F. (2023c). Fraz: A generic high-fidelity fixed-ratio lossy compression framework for scientific floating-point data. In IEEE IPDPS.
Collet, Y. (2025). lz4: Extremely fast compression algorithm. GitHub Repository. [link].
Elakkiya, S. and Thivya, K. (2022). Comprehensive review on lossy and lossless compression techniques. Journal of The Institution of Engineers (India): Series B, 103(3):1003–1012.
Fletcher, R. P., Du, X., and Fowler, P. J. (2009). Reverse time migration in tilted transversely isotropic (TTI) media. Geophysics, 74(6):WCA179–WCA187.
Knorr, F., Thoman, P., and Fahringer, T. (2021). ndzip-gpu: Efficient lossless compression of scientific floating-point data on gpus. In Proceedings SC21: International Conference for High Performance Computing, Networking, Storage and Analysis, pages 1–13.
Künas, C. A., Freytag, G., and Navaux, P. O. (2024). Enhancing reverse time migration simulations in hpc systems through i/o and computation overlapping. In Latin American High Performance Computing Conference. Springer.
Lindstrom, P., Hittinger, J., Diffenderfer, J., Fox, A., Osei-Kuffuor, D., and Banks, J. (2024). Zfp: A compressed array representation for numerical computations. The International Journal of High Performance Computing Applications.
Lindstrom, P. G. (2017). Fpzip. Technical report, Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States).
Liu, J., Tao, D., and Cappello, F. (2022). Understanding gpu-based lossy compression for extreme-scale cosmological simulations. Computing in Science & Engineering.
Liu, J., Tian, J., Wu, S., Di, S., Zhang, B., Huang, Y., Zhao, K., Li, G., Tao, D., Chen, Z., and Cappello, F. (2023). cusz: An efficient gpu-based error-bounded lossy compression framework for scientific data. arXiv preprint arXiv:2312.05492.
Liu, K., Di, S., Tao, D., and Cappello, F. (2020). Use cases of lossy compression for floating-point data in scientific data sets. Computing in Science & Engineering.
NVIDIA, C. (2025). The nvcomp library. [link]. Accessed: 2025-06-18.
Peñaranda, C., Reaño, C., and Silla, F. (2024). Hybrid-smash: a heterogeneous cpu-gpu compression library. IEEE Access, 12:32706–32723.
Qawasmeh, A., Hugues, M. R., Calandra, H., and Chapman, B. M. (2017). Performance portability in reverse time migration and seismic modelling via openacc. The International Journal of High Performance Computing Applications, 31(5):422–440.
Serpa, M. S., Pavan, P. J., Cruz, E. H., Machado, R. L., Panetta, J., Azambuja, A., Carissimi, A. S., and Navaux, P. O. (2021). Energy efficiency and portability of oil and gas simulations on multicore and graphics processing unit architectures. CCPE, 33(18):e6212.
Sruthi, S. and Meena, S. (2025). Comparative study of data compression algorithms: Zstandard, zlib & lz4. In 2025 International Conference on Data Engineering and Applications.
Zhao, K., Liu, J., Tao, D., and Cappello, F. (2022). Fz-gpu: A fast and high-ratio lossy compressor for scientific computing applications on gpus. In IEEE Cluster Conference.
Azami, N., Fallin, A., and Burtscher, M. (2025). Efficient lossless compression of scientific floating-point data on cpus and gpus. In Proceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 1, pages 395–409.
Barbosa, C. H. and Coutinho, A. L. (2022). Multi-gpu 3-d reverse time migration with minimum i/o. In Latin American High Performance Computing Conference, pages 160–173. Springer.
Chen, X., Tian, J., Beaver, I., Freeman, C., Yan, Y., Wang, J., and Tao, D. (2023a). Fcbench: Cross-domain benchmarking of lossless compression for floating-point data. Proceedings of the VLDB Endowment, 17(6):1418–1431.
Chen, X., Tian, J., Beaver, I., Freeman, C., Yan, Y., Wang, J., and Tao, D. (2024). Fcbench: Cross-domain benchmarking of lossless compression for floating-point data. Proceedings of the VLDB Endowment, 17(6):1418–1431.
Chen, Z., Di, S., Tao, D., and Cappello, F. (2023b). Beyond compression: A comprehensive evaluation of lossless floating-point compression. Journal of Parallel and Distributed Computing.
Chen, Z., Tao, D., and Cappello, F. (2023c). Fraz: A generic high-fidelity fixed-ratio lossy compression framework for scientific floating-point data. In IEEE IPDPS.
Collet, Y. (2025). lz4: Extremely fast compression algorithm. GitHub Repository. [link].
Elakkiya, S. and Thivya, K. (2022). Comprehensive review on lossy and lossless compression techniques. Journal of The Institution of Engineers (India): Series B, 103(3):1003–1012.
Fletcher, R. P., Du, X., and Fowler, P. J. (2009). Reverse time migration in tilted transversely isotropic (TTI) media. Geophysics, 74(6):WCA179–WCA187.
Knorr, F., Thoman, P., and Fahringer, T. (2021). ndzip-gpu: Efficient lossless compression of scientific floating-point data on gpus. In Proceedings SC21: International Conference for High Performance Computing, Networking, Storage and Analysis, pages 1–13.
Künas, C. A., Freytag, G., and Navaux, P. O. (2024). Enhancing reverse time migration simulations in hpc systems through i/o and computation overlapping. In Latin American High Performance Computing Conference. Springer.
Lindstrom, P., Hittinger, J., Diffenderfer, J., Fox, A., Osei-Kuffuor, D., and Banks, J. (2024). Zfp: A compressed array representation for numerical computations. The International Journal of High Performance Computing Applications.
Lindstrom, P. G. (2017). Fpzip. Technical report, Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States).
Liu, J., Tao, D., and Cappello, F. (2022). Understanding gpu-based lossy compression for extreme-scale cosmological simulations. Computing in Science & Engineering.
Liu, J., Tian, J., Wu, S., Di, S., Zhang, B., Huang, Y., Zhao, K., Li, G., Tao, D., Chen, Z., and Cappello, F. (2023). cusz: An efficient gpu-based error-bounded lossy compression framework for scientific data. arXiv preprint arXiv:2312.05492.
Liu, K., Di, S., Tao, D., and Cappello, F. (2020). Use cases of lossy compression for floating-point data in scientific data sets. Computing in Science & Engineering.
NVIDIA, C. (2025). The nvcomp library. [link]. Accessed: 2025-06-18.
Peñaranda, C., Reaño, C., and Silla, F. (2024). Hybrid-smash: a heterogeneous cpu-gpu compression library. IEEE Access, 12:32706–32723.
Qawasmeh, A., Hugues, M. R., Calandra, H., and Chapman, B. M. (2017). Performance portability in reverse time migration and seismic modelling via openacc. The International Journal of High Performance Computing Applications, 31(5):422–440.
Serpa, M. S., Pavan, P. J., Cruz, E. H., Machado, R. L., Panetta, J., Azambuja, A., Carissimi, A. S., and Navaux, P. O. (2021). Energy efficiency and portability of oil and gas simulations on multicore and graphics processing unit architectures. CCPE, 33(18):e6212.
Sruthi, S. and Meena, S. (2025). Comparative study of data compression algorithms: Zstandard, zlib & lz4. In 2025 International Conference on Data Engineering and Applications.
Zhao, K., Liu, J., Tao, D., and Cappello, F. (2022). Fz-gpu: A fast and high-ratio lossy compressor for scientific computing applications on gpus. In IEEE Cluster Conference.
Publicado
28/10/2025
Como Citar
KÜNAS, Cristiano A.; FREYTAG, Gabriel; ARAÚJO, Thiago; MACHADO, Rodrigo; MORALES, Bruno; SARDINHA, Alexandre; NAVAUX, Philippe O. A..
Compressão de Dados para Redução de E/S em Simulações Sísmicas de Alto Desempenho. In: SIMPÓSIO EM SISTEMAS COMPUTACIONAIS DE ALTO DESEMPENHO (SSCAD), 26. , 2025, Bonito/MS.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2025
.
p. 326-337.
DOI: https://doi.org/10.5753/sscad.2025.16726.
