Wave field reconstruction method three-dimensional pre-stack reverse time migration method based on FPGA multi-board card architecture

Abstract

The invention discloses a wave field reconstruction method three-dimensional pre-stack reverse time migration method based on an FPGA multi-board card framework, and relates to the technical field of computers, and the method comprises the steps: S1, preparing calculation data needed by forward modeling at a CPU side, and completing the programming of the calculation data through employing a programming language; s2, the CPU performs data interaction with the first FPGA board card, and the first FPGA board card outputs corresponding parameters to the second FPGA board card and the third FPGA board card; s3, the second FPGA board card calculates a forward modeling wave field, and after calculation is completed, an end signal is output to the first FPGA board card; s4, the first FPGA board card starts counting and outputs signals to the second FPGA board card and the third FPGA board card at the same time, and calculation of a forward wave field and a reverse wave field is started; s5, the second FPGA board and the third FPGA board output the forward wave field and the reverse wave field at the same moment to the first FPGA board; and S6, the first FPGA board completes calculation at all moments and outputs a final imaging result to the CPU. Unnecessary data cache transmission is reduced, consumption of computing resources is reduced, and computing efficiency is improved.

Description

technical field [0001] The invention relates to the field of computer technology, in particular to a three-dimensional pre-stack reverse time migration method based on a FPGA multi-board card architecture. Background technique [0002] In recent years, with the rapid development of computer technology and the substantial improvement of computing power, solutions for the large amount of computation and storage for reverse time migration have also been studied. Large-scale computing methods and architectures for industrial applications are also studied. The improvement of computing performance is mainly due to the improvement of the operating frequency of microprocessors, and Moore's Law dominates the pace of information technology development. The system heating problem brought about by the increase of the main frequency is becoming more and more prominent, and the power consumption of the system is constantly rising. These problems are also constantly restricted by physical...

AI Technical Summary

Problems solved by technology

[0004] However, the current high-density data collection and deep and ultra-deep imaging problems faced by oil exploration make the 3D-RTM calculation data volume huge and require high computing power

The current FPGA chip single card has limited hardware resources. When a computing board is used to calculate large-scale ultra-deep 3D-RTM, due to insufficient FPGA resources, a large amount of data needs to be transferred to the CPU. This repeated data reading is huge. Increased consumption of computing resources for data transfer

Images