In this paper we present methods of parallel programming on graphics processor unit (GPU). We describe and consider usefulness of NVIDIA's SDK CUDA (Compute Unified Device Architecture), which make GPGPU easier. Finally we try to estimate potential capacity of new graphics processors and also look closer for edge detection example program.
Published in:
Signal Processing Algorithms, Architectures, Arrangements, and Applications (SPA), 2008
Date of Conference: 25-27 Sept. 2008