HR: 1340h
AN: SF13A-0704    [Abstracts]
TI: gLucifer: Next-generation visualization framework
AU: Watson, G
EM: gdwat1@student.monash.edu
AF: Monash Cluster Computing, School of Mathematical Sciences, Monash University, Clayton, VIC 3800 Australia
AU: * Stegman, D R
EM: dave.stegman@sci.monash.edu
AF: Monash Cluster Computing, School of Mathematical Sciences, Monash University, Clayton, VIC 3800 Australia
AU: Duboz, C
EM: Cecile.Duboz@sci.monash.edu
AF: Monash Cluster Computing, School of Mathematical Sciences, Monash University, Clayton, VIC 3800 Australia
AU: Moresi, L
EM: louis.moresi@sci.monash.edu
AF: Monash Cluster Computing, School of Mathematical Sciences, Monash University, Clayton, VIC 3800 Australia
AB: We describe a general visualization framework, gLucifer, which is designed to complement general-purpose geodynamics programming frameworks such as StGermain, Snark, Underworld, and most any grid based computational models. The StGermain, Snark, and Underworld frameworks allow the rapid creation of scientific modeling applications which scale efficiently from small scale models on laptops and desktops to giant models on remote supercomputing or utilizing grid computing resources. gLucifer provides a visual analysis toolkit for multivariate and multi-dimensional data sets generated by the geodynamics frameworks, while simultaneously incorporating geological data. gLucifer provides both interactive and background rendering but concentrates on the latter for efficiently rendering frames of computationally intense, remotely running applications on distributed or clustered computers. gLucifer also provides a programming interface using XML which extends the geodynamics modeling frameworks to allow native visualization of objects such as meshes, particle swarms, embedded surfaces, scalar, vector and tensor fields on meshes and/or particle swarms with operations such as masking/colouring by node/particle or by ranges of values. The realizable goal of such an approach is for multi-dimensional data sets which are streaming out from a simulation to be dynamically probed, diagonosed and visualized using gLucifer during the time calculations are performed. This eliminates the significant time delay one typically experiences between having queued up a simulation to run and the point at which the results are post-processed and available for interpretation. In this way, gLucifer has a further advantage of removing one level of abstraction between the the end user and the simulation. gLucifer achieves all of this using open source, freely available libraries such as openGL, X11, VTK, and libfame (a movie encoder). Movies are now the standard output for such large, computationally intense simulations and we will present some example movies derived from thermal convection in a sphere, thermochemical convection in 3-D cartesian, and 3-D moving-mesh compressible/extensional simulations.
DE: 8147 Planetary interiors (5430, 5724)
DE: 5499 General or miscellaneous
SC: Special Focus: Advances in Data Acquisition, Management, Analysis and Display [SF]
MN: 2004 AGU Fall Meeting