Efficient Computation of Mandelbrot Set Generation with Compute Unified Device Architecture (CUDA)

Abstract

Compute Unified Device Architecture (CUDA) is one of the standards for interfaces in parallel programming implemented into NVIDIA's GPUs. GPU (Graphics Processing Unit) is a processor of many smaller, more specialized cores. With the implementation of CUDA, it is possible to find solutions to handle complex computational problems faster and more efficiently rather than the results of the computations performed by the CPU. Because creating images/videos requires many processes per pixel, experiments are applied to one example of computing problems related to them, which is fractal generation. The Mandelbrot set is one of the fractal generations that need a complex computational process to generate it. In this research, the execution of a Mandelbrot Set generation using the CPU and GPU (CUDA) will be done. Parallelization that occurs on CUDA’s Mandelbrot set generation, results in relatively faster execution time than sequential work on the CPU in certain cases.