Improving Random Walk Performance.
Ilya Safro, Paul Hovland, Jaewook Shin, Michelle Mills Strout · 2009
Abstract. Random walk simulation is employed in many experimental algorithmic applications. Efficient execution on modern computer archi-tectures demands that the random walk be implemented to exploit data locality for improving the cache performance. In this research, we demon-strate how different one-dimensional data reordering functionals can be used as a preprocessing step for speeding the random walk runtime. 1