Advances in GPU Research and Practice

Chapter 12

GPU sorting algorithms

M. Gopi    University of California, Irvine, CA, United States

Abstract

In this chapter, we briefly introduce the basics of graphics processing unit (GPU) architectures, CUDA programming, memory and computation models, and generic programming strategies to achieve high performance with GPU architectures. Using these models, we present GPU-based parallel implementations of popular sorting algorithms, including Bitonic Sort, Radix Sort, and Merge Sort. We also present a brief discussion about other sorting algorithms, such as Quicksort, Warpsort, and scalable sorting algorithms, along with a comparison between these sorting algorithms.

Keywords

GPU sorting algorithms; CUDA programming, memory, and computation ...