Accelerating experimental high-order spatial statistics calculations using GPUs

High-order spatial statistics have been widely used to describe the spatial phenomena in the field of geology science. Spatial statistics are subject to extremely heavy computational burden for large geostatistical models. To improve the computational efficiency, a parallel approach based on GPU (Graphics Processing Unit) is proposed for the calculation of high-order spatial statistics. The parallel scheme is achieved by utilizing a two-stage method to calculate the replicate of a moment for a given template simultaneously termed as the node-stage parallelism, and transform the spatial moments to cumulants for all lags of a template simultaneously termed as the template-stage parallelism. Also, a series of optimization strategies are proposed to take full advantage of the computational capabilities of GPUs, including the appropriate task allocation to the CUDA (Compute Unified Device Architecture) threads, proper organization of the GPU physical memory, and optimal improvement of the existed parallel routines. Tests are carried out on two training images to compare the performance of the GPU-based method with that of the serial implementation. Error analysis results indicate that the proposed parallel method can generate accurate cumulant maps, and the performance comparisons on various examples show that all the speedups for third-order, fourth-order and fifth-order cumulants calculation are over 17 times. A GPU-based method is initially proposed for high-order spatial statistics.Several optimizations are adopted for better performance.Experimental results show its superiorities in time consumption.

[1]  Naga K. Govindaraju,et al.  A Survey of General‐Purpose Computation on Graphics Hardware , 2007 .

[2]  Arie E. Kaufman,et al.  GPU Cluster for High Performance Computing , 2004, Proceedings of the ACM/IEEE SC2004 Conference.

[3]  Sheng Chen,et al.  Genetic algorithm optimization for blind channel identification with higher order cumulant fitting , 1997, IEEE Trans. Evol. Comput..

[4]  Roussos Dimitrakopoulos,et al.  Approximations of High-Order Spatial Statistics Through Decomposition , 2012 .

[5]  M. Rosenblatt,et al.  ASYMPTOTIC THEORY OF ESTIMATES OF kTH-ORDER SPECTRA. , 1967, Proceedings of the National Academy of Sciences of the United States of America.

[6]  G. Lazear.  Mixed-phase wavelet estimation using fourth-order cumulants , 1993 .

[7]  Jianting Zhang,et al.  Towards personal high-performance geospatial computing (HPC-G): perspectives and a case study , 2010, HPDGIS '10.

[8]  Pablo Fosalba,et al.  Gravitational Evolution of the Large-Scale Probability Density Distribution: The Edgeworth and Gamma Expansions , 2000 .

[9]  Peter J. Smith.  A Recursive Formulation of the Old Problem of Obtaining Moments from Cumulants and Vice Versa , 1995 .

[10]  Tao Huang,et al.  GPU-accelerated Direct Sampling method for multiple-point statistical simulation , 2013, Comput. Geosci..

[11]  Xianlin Ma,et al.  Modeling Conditional Distributions of Facies from Seismic Using Neural Nets , 2002 .

[12]  Pejman Tahmasebi,et al.  Accelerating geostatistical simulations using graphics processing units (GPU) , 2012, Comput. Geosci..

[13]  Georges Verly,et al.  The Practice of Sequential Gaussian Simulation , 2005 .

[14]  Tadeusz J. Ulrych,et al.  Simulated annealing wavelet estimation via fourth-order cumulant matching , 1996 .

[15]  Gang Wang,et al.  Efficient Parallel Lists Intersection and Index Compression Algorithms using Graphics Processing Units , 2011, Proc. VLDB Endow..

[16]  David Kirk,et al.  NVIDIA cuda software and gpu parallel computing architecture , 2007, ISMM '07.

[17]  R. Mohan Srivastava,et al.  Spatial continuity measures for probabilistic and deterministic geostatistics , 1988 .

[18]  E.Gaztanaga,et al.  Gravitational Evolution of the Large-Scale Probability Density Distribution: The Edgeworth&Gamma Expansions , 1999 .

[19]  Xing Peng,et al.  Seismic Wavelet Estimation Using High-Order Statistics and Chaos-Genetic Algorithm , 2012 .

[20]  Mario Chica-Olmo,et al.  Evaluation of Geostatistical Measures of Radiometric Spatial Variability for Lithologic Discrimination in Landsat TM Images , 1999 .

[21]  Andre G. Journel.  Beyond Covariance: The Advent of Multiple-Point Geostatistics , 2005 .

[22]  J. Vargas-Guzmán,et al.  The Kappa model of probability and higher-order rock sequences , 2011 .

[23]  Xu Wang,et al.  Computing prestack Kirchhoff time migration on general purpose GPU , 2011, Comput. Geosci..

[24]  S. Colombi,et al.  Large scale structure of the universe and cosmological perturbation theory , 2001, astro-ph/0112551.

[25]  Gérard Favier,et al.  Recursive estimation of fourth-order cumulants with application to identification , 1998, Signal Process..

[26]  R. Dimitrakopoulos,et al.  High-order Stochastic Simulation of Complex Spatially Distributed Natural Phenomena , 2010 .

[27]  Peter Bailey,et al.  Accelerating geoscience and engineering system simulations on graphics hardware , 2009, Comput. Geosci..

[28]  G.B. Giannakis,et al.  Cumulant-based order determination of non-Gaussian ARMA models , 1990, IEEE Trans. Acoust. Speech Signal Process..

[29]  Jens H. Krüger,et al.  GPGPU: general purpose computation on graphics hardware , 2004, SIGGRAPH '04.

[30]  Roussos Dimitrakopoulos,et al.  High-order Statistics of Spatial Random Fields: Exploring Spatial Cumulants for Modeling Complex Non-Gaussian and Non-linear Phenomena , 2009 .

[31]  Roussos G. Dimitrakopoulos,et al.  HOSIM: A high-order stochastic simulation algorithm for generating three-dimensional complex geological patterns , 2011, Comput. Geosci..

[32]  Gang Wang,et al.  Efficient lists intersection by CPU-GPU cooperative computing , 2010, 2010 IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum (IPDPSW).

[33]  Snehamoy Chatterjee,et al.  Geologic heterogeneity representation using high‐order spatial cumulants for subsurface flow and transport simulations , 2011 .

[34]  Julián M. Ortiz,et al.  Multiple-Point Geostatistical Simulation Based on Genetic Algorithms Implemented in a Shared-Memory Supercomputer , 2012 .

[35]  Edzer Pebesma,et al.  Spatial interpolation in massively parallel computing environments , 2011 .

[36]  G. Robertson,et al.  Spatial heterogeneity of soil respiration and related properties at the plant scale , 2000, Plant and Soil.

[37]  J. Chilès,et al.  Geostatistics: Modeling Spatial Uncertainty , 1999 .

[38]  Tao Huang,et al.  GPU-based SNESIM implementation for multiple-point statistical simulation , 2013, Comput. Geosci..

[39]  Qianyu Li,et al.  Using high-order cumulants to extrapolate spatially variant seismic wavelets , 2002 .

[40]  P. Goovaerts.  Geostatistical tools for characterizing the spatial variability of microbiological and physico-chemical soil properties , 1998, Biology and Fertility of Soils.

[41]  Dean S. Oliver.  Conditioning Channel Meanders to Well Observations , 2002 .

[42]  Kai Xie,et al.  GPU-Based Large Seismic Data Parallel Compression , 2013 .

[43]  Roussos G. Dimitrakopoulos,et al.  A new approach for geological pattern recognition using high-order spatial cumulants , 2010, Comput. Geosci..