Acceleration of Three-dimensional Direct Current Resistivity Modelling Using GPU Computing

Forward modelling algorithm is used both for modelling studies and for calculation of the theoretical data set in inversion algorithm. 3D inversion algorithms take long time due to data and parameter numbers. To reduce this time, some test studies have been made in forward modelling algorithm using the advantages of developing graphics card technology. Forward modelling equation is an elliptic type partial differential equation which is called Poisson equation. This equation is solved by finite-difference method and the results are obtained with conjugate gradient method. The conjugate gradient CUDA algorithm is written in C language. The speed-up for calculation time is showed for synthetic model studies.