Adaptive Performance-Constrained In Situ Visualization of Atmospheric Simulations

Abstract : While many parallel visualization tools now provide in situ visualization capabilities, the trend has been to feed such tools with large amounts of unprocessed output data and let them render everything at the highest possible resolution. This leads to an increased run time of simulations that still have to complete within a fixed-length job allocation. In this paper, we tackle the challenge of enabling in situ visualization under performance constraints. Our approach shuffles data across processes according to its content and filters out part of it in order to feed a visualization pipeline with only a reorganized subset of the data produced by the simulation. Our framework leverages fast, generic evaluation procedures to score blocks of data, using information theory, statistics, and linear algebra. It monitors its own performance and adapts dynamically to achieve appropriate visual fidelity within predefined performance constraints. Experiments on the Blue Waters supercomputer with the CM1 simulation show that our approach enables a 5× speedup with respect to the initial visualization pipeline, and is able to meet performance constraints.
Type de document :
Communication dans un congrès
Cluster 2016 - The IEEE 2016 International Conference on Cluster Computing, 2016, Taipei, Taiwan. 2016
Liste complète des métadonnées

Littérature citée [24 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01351919
Contributeur : Gabriel Antoniu <>
Soumis le : mercredi 31 août 2016 - 15:42:01
Dernière modification le : mercredi 6 septembre 2017 - 11:25:52
Document(s) archivé(s) le : vendredi 2 décembre 2016 - 06:30:51

Fichier

SmartViz-final-1.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01351919, version 1

Citation

Matthieu Dorier, Robert Sisneros, Leonardo Bautista-Gomez, Tom Peterka, Leigh Orf, et al.. Adaptive Performance-Constrained In Situ Visualization of Atmospheric Simulations. Cluster 2016 - The IEEE 2016 International Conference on Cluster Computing, 2016, Taipei, Taiwan. 2016. 〈hal-01351919〉

Partager

Métriques

Consultations de
la notice

739

Téléchargements du document

78