Implementing high performance voltammetry simulation using the implicit parallel algorithm

Abstract

The numerical simulation of voltammetry and amperometry at a microdisc electrode is implemented with a parallel algorithm using GPU (CUDA in C++) and adapting the implicit finite difference approximation with the ADI method. This leads to high performance simulation of the current response whilst maintaining the high standard of accuracy required for such electrochemical systems. We gained a maximum speed-up of 20 times with a standard graphic card specifications (Quadro K4000) in comparison to a single core of 3.2GHz CPU. The simplified allocation and data flow allows an accessible code which produces a starting point for researchers to manipulate the code to study new problems in electrochemical and surface reaction systems that require high performance simulation.