Using Compute Unified Device Architecture (CUDA) in Parallelizing Different Digital Image Processing Techniques
ICS University of the Philippines Los Bano
ICS University of the Philippines Los Bano, CMSC 190, 2012
Graphics Processing Units (GPUs) have been conventionally used in the acceleration of 2D, 3D graphics and video rendering. Because of its performance and capability, the GPU has evolved into a highly parallel programmable processor that specializes in memory bandwith utilization and intensive computation. For operations involving graphics, GPUs offer a lot of gigaflops of processing prowess. The programmability and competency of GPU in the field of general-purpose computing is exemplified through the implementation of image processing techniques using Compute Unified Device Architecture (CUDA) which makes image processing be implemented with a high-speed performance.
May 25, 2012 by hgpu