Difference between revisions of "SWE"

From Sccswiki
Jump to navigation Jump to search
Line 10: Line 10:
 
* SWE was used within the [http://icsweb.inf.unisi.ch/cms/index.php/component/content/article/12-news/95-summerschool2013.html CSCS-USI Summer School on Computer Simulations in Science and Engineering] (July 8-19, 2013) for demonstrations and tutorials regarding parallelisation.
 
* SWE was used within the [http://icsweb.inf.unisi.ch/cms/index.php/component/content/article/12-news/95-summerschool2013.html CSCS-USI Summer School on Computer Simulations in Science and Engineering] (July 8-19, 2013) for demonstrations and tutorials regarding parallelisation.
 
* SWE was used for a course on GPU computing with NVIDIA CUDA during the [http://www.mac.tum.de/g2s3/ Gene Golub SIAM Summer School 2012] on Simulation and Supercomputing in the Geosciences
 
* SWE was used for a course on GPU computing with NVIDIA CUDA during the [http://www.mac.tum.de/g2s3/ Gene Golub SIAM Summer School 2012] on Simulation and Supercomputing in the Geosciences
** see the respective [http://www.mac.tum.de/g2s3/swe_g2s3_2012.pdf slides on SWE]
+
** [http://www.mac.tum.de/g2s3/swe_g2s3_2012.pdf slides on SWE and on its CUDA implementation] are available
  
 
= Availability of the Code =
 
= Availability of the Code =

Revision as of 09:45, 17 July 2013

An Education-Oriented Code for Parallel Tsunami Simulation

SWE is an education-oriented code that implements simple Finite Volumes models that solve the shallow water equations - in a problem setting as it would be used for tsunami simulation. SWE has a modular design that allows parallelisation using different programming paradigms, such as MPI, OpenMP, or CUDA (further tests were done with Intel TBB/ArBB and OpenCL).

Documentation

  • An article on the design and intent of SWE was accepted for the 11th International Symposium on Parallel and Distributed Computing - ISPDC 2012 - it will be published by IEEE in the respective conference proceeding (late 2012). You can download a preprint of the paper (as PDF).
  • A doxygen documentation of the code is availble (note that the linked version here is not necessarily the latest version available from github).

Course Material

Availability of the Code

SWE is made available under the GPL license. The latest version can be downloaded via git. We are currently doing a lot of changes (to be used during the Gene Golub SIAM Summer School 2012, and plan to publish a complete version in August/September 2012).

Simulation Examples

Tohoku surface 46min.png

Tohoku both 224min.png