Accelerating Action Potential Generation Using GPU Implementation of a Resonant Model of a Cell

Computational modelling of bioelectric phenomena in cardiac tissue has made significant contributions to our understanding of the electrical activity of the human heart. However, extending these simulations over large temporal and spatial ranges is an inherently computationally expensive task. Motivated by the apparent advantages of GPU over CPU in many areas of science, we have developed a new Resonant Model (RM) of an AP of a biological cell amenable to parallel computing. In this study, we have investigated the potential performance benefits of the implementation of the RM on a GPU over the sequential execution on a CPU. In the proposed implementation, all threads in the same GPU warp share data using a warp-shuffle operation, which eliminates the access of global or shared memory. By exploiting the memory model and computational strengths of GPU, the obtained performance of the RM on the GPU is far superior to that of the CPU. On the GPU, the RM demonstrate linear or sub-linear growth in the execution times with the increase in the number of cells. The speedup of the computations achieved on the GPU may, in the near future, facilitate the patient-specific electrophysiological studies and treatment planning.