Scalable Unstructured Mesh Computation

Lori Freitag, Carl Ollivier-Gooch, Mark Jones, and
Paul Plassmann


Welcome to the SUMAA3D Project, a collaborative research effort between scientists at Argonne National Laboratory and The University of Tennessee.

This project was designed to enable the solution of large-scale applications on massively parallel machines. To date, many applications have been unable to take advantage of these machines because of a lack of publicly available software tools for unstructured mesh computation. The SUMAA3D project will rectify this situation by providing highly efficient parallel algorithms and software for the fundamental tasks of unstructured mesh computation.

In this demonstration, you will find a description of the project as well as several movies illustrating parallel computation on unstructured meshes. In addition, we have included detailed descriptions of the software packages already publicly available, applications which have benefited from this software, and our most recent work investigating the utility of virtual environment visualization techniques.


The SUMAA3D Project

An Overview -- The goals of SUMAA3D.
An Example -- Refinement of shell elements for a 3-dimensional geometry.
Accomplishments -- What is currently available.

Scalable Software

Mesh Smoothing Algorithms -- Parallel software for smoothing valid meshes.
Adaptive Mesh Refinement Algorithms -- Parallel software for adaptive, conforming meshes.
Linear System Solution using BlockSolve95 -- An efficient parallel solver for symmetric, sparse systems.


Large-Scale Applications and Virtual Environments

High-Temperature Superconductors -- A look at three-dimensional vortex structures.
Piezoelectric Crystals -- Finding vibrational modes of the crystals.
Commercial Combustion Systems -- Interactive supercomputing using virtual reality!
Interactive Mesh Refinement -- Refining a mesh from the CAVE.

Additional Information

Related Papers -- Electronic versions of recent publications.


Project Scientists


freitag@mcs.anl.gov / gooch@mcs.anl.gov / jones@cs.utk.edu / plassman@mcs.anl.gov
Argonne National Laboratory / The University of Tennessee

View the usage statistics from the previous day.