Parallel Seismic Modeling Based on OpenMP+AVX and Optimization Strategy

Abstract

This paper describes parallel simulation of the memory/computation-intensive acoustic wave equation with CPU template buffer optimization. Considering the 8-core CPU shared storage platform as an example, we obtain a one-time speed-up ratio of 6.7× compared with the serial program by using a coarse-grained OpenMP parallel scheme. Then, data is vectorized on the template buffer using the single instruction-multiple data (SIMD) technique to further exploit the computing potential of the CPUs. We apply an 8-channel parallel vector to simulate seismic wavefields with the 256-bit advanced vector extensions (AVX) instruction set. This increases the computing bandwidth, thus eliminating a significant volume of the computing instructions and obtaining a secondary speed-up ratio of 3–7×. In addition, we use 32-byte data alignment, shortest data direction vectorization, and loop tiling optimization algorithm to achieve faster program execution. Finally, we analyze the factors affecting the secondary speed-up of AVX through three-dimensional modeling experiments with the salt model. The results indicate that the memory, cache, and register can better cooperate with each other and the speed-up is increased by optimizing the AVX algorithm.

Keywords

forward modeling / OpenMP / AVX / parallel optimization

References

[1] Agulleiro J. I., Fernandez J. J. Tuning the Cache Memory Usage in Tomographic Reconstruction on Standard Computers with Advanced Vector Extensions (AVX). Data in Brief, 2015, 3: 16-20.

[2] Bian A. F., Zou Z. H., Zhou H. W., . Evaluation of Multi-Scale Full Waveform Inversion with Marine Vertical Cable Data. Journal of Earth Science, 2015, 26 4 481

[3] Calandra H., Bothorel F., Vezolle P. A Massively Parallel Implementation of the Common Azimuth Pre-Stack Depth Migration. IBM Journal of Research and Development, 2008, 52(1/2): 83-91.

[4] Etgen J. T., O’Brien M. J. Computational Methods for Large-Scale 3D Acoustic Finite-Difference Modeling: A Tutorial. Geophysics, 2007, 72(5): SM223-SM230.

[5] Francés J., Bleda S., Márquez A., . Performance Analysis of SSE and AVX Instructions in Multi-Core CPUs and GPU Computing on FDTD Scheme for Solid and Fluid Vibration Problems. The Journal of Supercomputing, 2014, 70 2 514

[6] Gregory K., Miller A. C++ AMP: Accelerated Massive Parallelism with Micorsoft Visual C++. Microsoft Press, Redmond, 2012, 127-170.

[7] Gokhberg A., Fichtner A. Full-Waveform Inversion on Heterogeneous HPC Systems. Computers & Geosciences, 2016, 89: 260-268.

[8] Huang W., Zhou H. W. Least-Squares Seismic Inversion with Stochastic Conjugate Gradient Method. Journal of Earth Science, 2015, 26 4 463

[9] Jayaseelan R., Liu H., Mitra T. Exploiting Forwarding to Improve Data Bandwidth of Instruction-Set Extensions, 2006, 24-28.

[10] Mojica O. F., Bassrei A. Generalized Cross-Validation and Regínska’s Methods for Choosing the Regularization Parameter in 3D Gravity Inversion of Basement Relief—A Hybrid MPI/OpenMP Parallel Algorithm, 2015, 3-6.

[11] de Janeiro Ronn F. Cache-Oblivious Searching and Sorting: [Dissertation], 2003, 7-36.

[12] Souza P., Borges L., Andreolli C., . OpenVec Portable SIMD Intrinsics, 2015, 13-16.

[13] Stock K. A. Vectorization and Register Reuse in High Performance Computing: [Dissertation], 2014, 16-21.

[14] Zhang L. Z., Du Y. X., Wu D. C. GPU-Accelerated FDTD Simulation of Human Tissue Using C++ AMP, 2015, 22-26.

[15] Zhou M. H., Symes W. W. Wave Equation Based Stencil Optimizations on Multi-Core CPU, 2014, 26-31.