Techniques for efficient DCT/IDCT implementation on generic GPU

@inproceedings{citeulike:1885737, abstract = {The emergence of programmable graphics processing units (GPU) has led to increasing interest in off-loading numerically intensive computations on to graphics hardware. DCT/IDCT is widely adopted in modern image/video compression standards and is usually one of the most computationally expensive parts. We present several techniques for efficient implementation of DCT/IDCT on generic programmable GPU, using direct matrix multiplication. Our experimental results demonstrate that the speed of IDCT on a GPU using the proposed techniques can well exceed that on a CPU with MMX optimization.}, author = {Fang, Bo and Shen, Guobin and Li, Shipeng and Chen, Huifang }, booktitle = {Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on}, citeulike-article-id = {1885737}, journal = {Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on}, keywords = {computing, gpu}, pages = {1126--1129 Vol. 2}, posted-at = {2007-11-08 19:00:36}, priority = {2}, title = {Techniques for efficient DCT/IDCT implementation on generic GPU}, url = {http://ieeexplore.ieee.org/xpls/abs\_all.jsp?arnumber=1464791}, year = {2005} }

TY - CONF ID - citeulike:1885737 L3 - citeulike-article-id:1885737 TI - Techniques for efficient DCT/IDCT implementation on generic GPU T2 - Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on JF - Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on SP - 1126 EP - 1129 Vol. 2 N2 - The emergence of programmable graphics processing units (GPU) has led to increasing interest in off-loading numerically intensive computations on to graphics hardware. DCT/IDCT is widely adopted in modern image/video compression standards and is usually one of the most computationally expensive parts. We present several techniques for efficient implementation of DCT/IDCT on generic programmable GPU, using direct matrix multiplication. Our experimental results demonstrate that the speed of IDCT on a GPU using the proposed techniques can well exceed that on a CPU with MMX optimization. KW - computing KW - gpu AU - Fang, Bo AU - Shen, Guobin AU - Li, Shipeng AU - Chen, Huifang PY - 2005/// UR - http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1464791 ER -