Performance modeling in action: Performance prediction of a Cray XT4 system during upgrade

Author

Barker, Kevin J. ; Davis, Kei ; Kerbyson, Darren J.

Author_Institution

Performance & Archit. Lab. (PAL), Los Alamos Nat. Lab., Los Alamos, NM, USA

fYear

2009

fDate

23-29 May 2009

Firstpage

1

Lastpage

8

Abstract

We present predictive performance models of two of the petascale applications, S3D and GTC, from the DOE Office of Science workload. We outline the development of these models and demonstrate their validation on an Opteron/Infiniband cluster and the pre-upgrade ORNL Jaguar system (Cray XT3/XT4). Given the high accuracy of the full application models, we predict the performance of the Jaguar system after the upgrade of its nodes, and subsequently compare this to the actual performance of the upgraded system. We then analyze the performance of the system based on the models to quantify bottlenecks and potential optimizations. Finally, the models are used to quantify the benefits of alternative node allocation strategies.

Keywords

Cray computers; computer network performance evaluation; parallel machines; workstation clusters; Cray XT3 system; Cray XT4 system; DOE; GTC; Office of Science; Opteron-Infiniband cluster; S3D; node allocation strategies; petascale applications; predictive performance models; preupgrade ORNL Jaguar system; Delay; Hardware; Laboratories; Measurement; Network topology; Performance analysis; Performance evaluation; Predictive models; Routing; US Department of Energy;

fLanguage

English

Publisher

ieee

Conference_Titel

Parallel & Distributed Processing, 2009. IPDPS 2009. IEEE International Symposium on

Conference_Location

Rome

ISSN

1530-2075

Print_ISBN

978-1-4244-3751-1

Electronic_ISBN

1530-2075

Type

conf

DOI

10.1109/IPDPS.2009.5161098

Filename

5161098