PROJECT TITLE :

Implementation of the DWT in a GPU through a Register-based Strategy

ABSTRACT:

The unleash of the CUDA Kepler architecture in March 2012 has provided Nvidia GPUs with a bigger register memory area and instructions for the communication of registers among threads. This facilitates a new programming strategy that utilizes registers for information sharing and reusing in detriment of the shared memory. Such a programming strategy will significantly improve the performance of applications that reuse data heavily. This paper presents a register-based mostly implementation of the Discrete Wavelet Transform (DWT), the prevailing information decorrelation technique in the field of image coding. Experimental results indicate that the proposed technique is, a minimum of, four times faster than the best GPU implementation of the DWT found in the literature. Furthermore, theoretical analysis coincide with experimental tests in proving that the execution times achieved by the proposed implementation are close to the GPU’s performance limits.