Alternative Computing Platforms to Implement the Seismic Migration

Abstract

Introducion: High performance computing (HPC) fails to satisfy the computational requirements of Seismic Migration (SM) algorithms, due to problems related to computing, the data transference and management ([4]). Therefore, the execution time of these algorithms in state-of-the-art clusters still remain in the order of months [1]. Since SM is one of the standard data processing techniques used in seismic exploration, because it generates an accurate image of the subsurface, oil companies are particularly interested in reducing execution times of SM algorithms. Computational migration needed for large datasets acquired today is extremely demanding, and therefore the computation requires CPU clusters. The performance of CPU clusters had been duplicating each 24 months until 2000 (satisfying the SM demands), but since 2001 this technology stop accelerating due to three technological limitations known as Power Wall, Memory Wall and IPL Wall [10, 23]. This encouraged experts all over the world to find new computing alternatives. The devices that have been highlighted as a base for the alternative computing platforms are FPGAs and GPGPUs. These technologies are subject of major research in HPC since they have a better perspective of computing [10, 14]. Different implementations of SM algorithms have been developed using those alternatives platforms [4, 12, 17, 21, 22, 29]. Results show a reduction in the running times of SM algorithms, leading to combine these alternative platforms with traditional CPU clusters in order to get a promising future in HPC for seismic exploration. This chapter gives an overview of the HPC with an historical perspective, emphasizing in the technological aspects related to the SM. In the section two we will show the seismic migration evolution together with the technology, section three summarizes the most important aspects of the CPU operation in order to understand the three technological walls, section four presents the use of GPUs as a new HPC platform to implement the SM, section five shows the use of FPGAs as another HPC platform, section six discusses the advantages and disadvantages of both technologies and, finally, the chapter is closed with the acknowledgments. At the end of this chapter,it is expected that the reader have enough background to compare platform specifications and be able to choose the most suitable platform for the implementation of the SM.