Optimization of linked list prefix computations on multithreaded GPUs using CUDA

Abstract

We present a number of optimization techniques to compute prefix sums on linked lists and implement them on multithreaded GPUs using CUDA. Prefix computations on linked structures involve in general highly irregular fine grain memory accesses that are typical of many computations on linked lists, trees, and graphs. While the current generation of GPUs… (More)
DOI: 10.1142/S0129626412500120

Topics

11 Figures and Tables

Cite this paper

@article{Wei2010OptimizationOL, title={Optimization of linked list prefix computations on multithreaded GPUs using CUDA}, author={Zheng Wei and Joseph J{\'a}J{\'a}}, journal={2010 IEEE International Symposium on Parallel & Distributed Processing (IPDPS)}, year={2010}, pages={1-8} }