A Performance Study of Applications in the Australian Community Climate and Earth System Simulator - Inria - Institut national de recherche en sciences et technologies du numérique
Communication Dans Un Congrès Année : 2015

A Performance Study of Applications in the Australian Community Climate and Earth System Simulator

Mark Cheeseman
  • Fonction : Auteur
  • PersonId : 983334
Ben Evans
  • Fonction : Auteur
  • PersonId : 983335
Dale Roberts
  • Fonction : Auteur
  • PersonId : 983333
Marshall Ward
  • Fonction : Auteur
  • PersonId : 983331

Résumé

A 3-year investigation is underway into the performance of applications used in the Australian Community Climate and Earth System Simulator on the petascale supercomputer Raijin hosted at the National Computational Infrastructure. Several applications have been identified as candidates for this investigation including the UK MetOffice’s Unified Model (UM) atmospheric model and Princeton University’s Modular Ocean Model (MOM). In this paper we present initial results of the investigation of the performance and scalability of UM and MOM on Raijin. We also present initial results of a performance study on the data assimilation package (VAR) developed by the UK MetOffice and used by the Australian Bureau of Meteorology in its operational weather forecasting suite. Further investigation and optimization is envisioned for each application investigated and will be discussed.
Fichier principal
Vignette du fichier
978-3-319-15994-2_60_Chapter.pdf (414.38 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01328611 , version 1 (08-06-2016)

Licence

Identifiants

Citer

Mark Cheeseman, Ben Evans, Dale Roberts, Marshall Ward. A Performance Study of Applications in the Australian Community Climate and Earth System Simulator. 11th International Symposium on Environmental Software Systems (ISESS), Mar 2015, Melbourne, Australia. pp.589-598, ⟨10.1007/978-3-319-15994-2_60⟩. ⟨hal-01328611⟩
95 Consultations
103 Téléchargements

Altmetric

Partager

More