DMCA
Parallel white noise generation on a gpu via cryptographic hash (2008)
| Venue: | In Proceedings of SI3D |
| Citations: | 30 - 2 self |
BibTeX
@INPROCEEDINGS{Tzeng08parallelwhite,
author = {Stanley Tzeng and Li-yi Wei},
title = {Parallel white noise generation on a gpu via cryptographic hash},
booktitle = {In Proceedings of SI3D},
year = {2008},
pages = {79--87}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
A good random number generator is essential for many graphics applications. As more such applications move onto parallel pro-cessing, it is vital that a good parallel random number generator be used. Unfortunately, most random number generators today are still sequential, exposing performance bottlenecks and denying random accessibility for parallel computations. Furthermore, popular paral-lel random number generators are still based off sequential methods and can exhibit statistical bias. In this paper, we propose a random number generator that maps well onto a parallel processor while possessing white noise distribu-tion. Our generator is based on cryptographic hash functions whose statistical robustness has been examined under heavy scrutiny by cryptologists. We implement our generator as a GPU pixel pro-gram, allowing us to compute random numbers in parallel just like ordinary texture fetches: given a texture coordinate per pixel, in-stead of returning a texel as in ordinary texture fetches, our pixel program computes a random noise value based on this given tex-ture coordinate. We demonstrate that our approach features the best quality, speed, and random accessibility for graphics applications.
Keyphrases
parallel white noise generation cryptographic hash random accessibility ordinary texture fetch good parallel random number generator popular paral-lel random number generator gpu pixel pro-gram tex-ture coordinate random number generator texture coordinate pixel program parallel computation random number generator today good random number generator parallel processor graphic application random noise value statistical robustness random number many graphic application statistical bias performance bottleneck parallel pro-cessing heavy scrutiny white noise distribu-tion sequential method cryptographic hash function
