Overlapping Computation and Writing in the Fletcher Method with MPI

• Rodrigo C. Machado UFRGS
• Arthur F. Lorenzon UFRGS
• Philippe O. A. Navaux UFRGS

DOI: https://doi.org/10.5753/sscad.2024.244792

Abstract

The oil and gas industry needs to drill sites to locate new reservoirs of oil and gas. To make more informed decisions on where to drill, the industry utilizes seismic surveys, which involve sending and capturing seismic waves in the ocean. The Fletcher method is an algorithm that uses the collected data to run simulations. In this process, images of the ocean subsoil are produced and must be stored for later analysis. Writing these images to a storage unit accounts for a significant portion of the application’s total time. To optimize the Fletcher method, we split and overlapped the computation and writing parts of the application using MPI. With our contribution, we achieved our best results with a speedup of 1.17x for executions on HD and 1.20x for SSD.

References

ANP (2023). Como funciona o processo de exploração e produção de petróleo e gás natural no brasil. Acesso em: 22 jul. 2024.

Chowdhury, M. K. H., Tang, H., Bez, J. L., Bangalore, P. V., and Byna, S. (2023). Efficient asynchronous i/o with request merging. In Proceedings of the IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW), pages 628–636.

Corbett, P., Feitelson, D., Fineberg, S., Hsu, Y., Nitzberg, B., Prost, J.-P., Snirt, M., Traversat, B., and Wong, P. (1996). Overview of the MPI-IO Parallel I/O Interface, pages 127–146. Springer US, Boston, MA.

Fletcher, R. P., Du, X., and Fowler, P. J. (2009). Reverse time migration in tilted transversely isotropic (TTI) media. Geophysics, 74(6):179–187.

Forum, T. M. (2003). MPI-2: Extensions to the MPI Standard. Acesso em: 11 ago. 2024.

Open MPI (2018). Open MPI: Open Source High Performance Computing. Acessado em 16/02/2024.

Patrick, C. M., Son, S., and Kandemir, M. (2008). Comparative evaluation of overlap strategies with study of i/o overlap in mpi-io. SIGOPS Oper. Syst. Rev., 42(6):43–49.

Song, H.-J. and Lee, Y.-H. (2013). A study on the disk performance comparison. International Journal of Multimedia and Ubiquitous Engineering, 8.

Tang, H., Koziol, Q., Ravi, J., and Byna, S. (2022). Transparent asynchronous parallel i/o using background threads. IEEE Transactions on Parallel and Distributed Systems, 33(4):891–902.

Tipu, A. J. S., Conbhuí, P. , and Howley, E. (2022). Seismic data io and sorting optimization in hpc through anns prediction based auto-tuning for exseisdat. Neural Computing and Applications, 35:5855–5888.

Tsujita, Y., Yoshinaga, K., Hori, A., Sato, M., Namiki, M., and Ishikawa, Y. (2014). Multithreaded two-phase i/o: Improving collective mpi-io performance on a lustre file system. In Proceedings of the 22nd Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP), pages 232–235.