Degree Type

Thesis

Date of Award

2016

Degree Name

Master of Science

Department

Physics and Astronomy

Major

High Energy Physics

First Advisor

Chunhui Chen

Abstract

The advent and proliferation of modern graphical processing units (GPUs) containing hundreds to thousands of cores opens up the new possibility of rewriting programs to execute on GPUs and acheive significant gains in speed over the original implementation. We examine two algorithms taken from high-energy physics: topological clustering, and the anti-k_{T} jet finding algorithm. While no performance gains were attained for topological clustering, the execution time of the anti-k_{T} algorithm is nearly halved.

DOI

https://doi.org/10.31274/etd-180810-5386

Copyright Owner

Matthew Joseph Lockner

Language

en

File Format

application/pdf

File Size

54 pages

antikt.zip (358 kB)

Share

COinS