Recently graphics processors have been developed that allow programmers to execute small programs on their specialized rendering hardware. Although the original intent of these customizable rendering programs was to manipulate textures and images for gaming purposes, they are becoming more capable at performing general purpose algorithms. This thesis examines different algorithms implemented on a Graphics Processing Unit (GPU). These algorithms include: image processing, cardiac tis­sue simulation using cellular automata and sparse matrix operations. It was found that all algorithms showed speed improvements using the GPU over the CPU. The greatest speed improvements found were sparse matrix operations, where the GPU performed operations up to 14 times faster than the CPU.