Performance evaluation of explicit finite difference algorithms with varying amounts of computational and memory intensity


Jammy, Satya, Jacobs, Christian and Sandham, Neil (2016) Performance evaluation of explicit finite difference algorithms with varying amounts of computational and memory intensity Journal of Computational Science, pp. 1-15.

Download

[img] PDF Performance evaluation of explicit finite difference algorithms with varying amounts of computational and memory intensity.pdf - Accepted Manuscript
Restricted to Repository staff only until 27 January 2018.
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (478kB)

Description/Abstract

Future architectures designed to deliver exascale performance motivate the need for novel algorithmic changes in order to fully exploit their capabilities. In this paper, the performance of several numerical algorithms, characterised by varying degrees of memory and computational intensity, are evaluated in the context of finite difference methods for fluid dynamics problems. It is shown that, by storing some of the evaluated derivatives as single thread- or process-local variables in memory, or recomputing the derivatives on-the-fly, a speed-up of ~2 can be obtained compared to traditional algorithms that store all derivatives in global arrays.

Item Type: Article
ISSNs: 1877-7503 (print)
Related URLs:
Organisations: Aerodynamics & Flight Mechanics Group
ePrint ID: 402148
Date :
Date Event
27 October 2016Accepted/In Press
Date Deposited: 02 Nov 2016 13:58
Last Modified: 17 Apr 2017 01:11
Further Information:Google Scholar
URI: http://eprints.soton.ac.uk/id/eprint/402148

Actions (login required)

View Item View Item