CENTER Carnegie Mellon UniversityCarnegie Mellon Computer Science DepartmentSchool of Computer Science
Related Activities
Outreach Roadshow
The Terascale Simulation Tools and Technologies Center: Creating Interoperable Meshing and Discretization Components

Lori Freitag.


The Terascale Simulation Tools and Technologies Center (TSTT) is a multi-institution collaboration whose primary goal is to enable high-fidelity physics calculations using multiple meshing and discretization solution strategies within a single simulation. To accomplish this goal, we are working toward the development of interoperable meshing and discretization techniques by defining common interfaces for accessing and modifying mesh and geometry information. The interface design is driven by application requirements and the need for intuitive, easy-to-use interfaces at multiple levels of sophistication. Thus, we will provide both high-level abstractions (e.g., representations of an entire complex mesh structure, and operations on that mesh) appropriate for new application development and low-level access functions (e.g., approximations of derivatives at a single point on a mesh) appropriate for incremental insertion of new technologies into existing applications. Our initial work focuses on query interfaces for accessing information pertaining to low-level mesh objects such as vertices, edges, faces, and regions, for accessing coordinate and adjacency information, and for setting and retrieving user-defined tag information on mesh entities. Discussions are under way to determine interfaces for mesh services, canonical ordering of entities, and query interfaces for distributed meshes in a parallel computing environment.

In this talk I will give an overview of the TSTT center activities and describe the challenges associated with development of common interfaces including maintaining flexibility, performance, and language independence. I describe our solution to these challenges and discuss the use of our interfaces with various TSTT tools. Finally, I will comment on the risks associated with common interface development work and some lessons learned during the past year.


This material is based upon work supported by National Science Foundation under Grant No. 0122581.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the
National Science Foundation