A correctly beating heart is important to ensure adequate circulation of blood throughout the body. Normal heart rhythm is produced by the orchestrated conduction of electrical signals throughout the heart. Cardiac electrical activity is the resulted function of a series of complex biochemical-mechanical reactions, which involves transportation and bio-distribution of ionic flows through a variety of biological ion channels. Cardiac arrhythmias are caused by the direct alteration of ion channel activity that results in changes in the AP waveform. In this work, we developed a whole-heart simulation model with the use of massive parallel computing with GPGPU and OpenGL. The simulation algorithm was implemented under several different versions for the purpose of comparisons, including one conventional CPU version and two GPU versions based on Nvidia CUDA platform. OpenGL was utilized for the visualization / interaction platform because it is open source, light weight and universally supported by various operating systems. The experimental results show that the GPU-based simulation outperforms the conventional CPU-based approach and significantly improves the speed of simulation. By adopting modern computer architecture, this present investigation enables real-time simulation and visualization of electrical excitation and conduction in the large and complicated 3D geometry of a real-world human heart.