並列な整列アルゴリズム

最近GPUとCUDAを使用して並列に計算するプログラムを学んでいます。 I recently encountered an algorithm called scan, which, given an array of elements, replaces the ith element by the sum of elements 0 to i. Applications of this supposedly include sorting and histogram generation, but I am not certain of how this application is carried out. I am looking at the radix sort discussion in "GPU Gems 3" to gain further understanding.