AuthorsT. Gillberg
TitleDeveloping and Implementing Parallel Algorithms for Fold Simulations in 3D
AfilliationCenter for Biomedical Computing (SFF), Scientific Computing
Project(s)Center for Biomedical Computing (SFF)
StatusPublished
Publication TypeTalks, contributed
Year of Publication2013
Location of TalkPresented during a workshop at Simula: Using the GPU for Scientific Computing: coding concepts and a look at applications in structural geology http://simula.no/intranet/news/using-the-gpu-for-scie\ ntific-computing-coding-concepts-and-a-look-at-appl
KeywordsWorkshop
Abstract

The Compound Earth Simulator is an exploration software developed by Kalkulo for Statoil. In the Compound methodology, large folded geological structures are simulated repeatedly as geological models are investigated. Folded layers are modeled as positions of a front moving through a geological, and are computed using first-arrival-time equations. The numerical simulation times must be low for the software to function at interactive times, which is a challenge in 3D as the number of unknown nodes increases rapidly. This talk presents a new set of algorithms for such simulations, specifically developed for massively parallel compute architectures such as GPUs. A couple of optimization approaches and their effects are presented for the GPU implementation. Finally, some early result from a multicore CPU and a GPU implementation will be given.

Notes

Presented during a workshop at Simula: Using the GPU for Scientific Computing: coding concepts and a look at applications in structural geology http://simula.no/intranet/news/using-the-gpu-for-scientific-computing-co...

Citation KeySimula.simula.2004