NGS: Mesh Generation and Optimistic Computation on the Grid

College Of William And Mary, Williamsburg VA

Investigators

Abstract

EIA-0203974 Chrisochoides, Nikos P. College of William and Mary Mesh Generation and Optimistic Computation on the Grid . The goal of this proposal is to investigate exactly these issues using an identified application as a test case. Mesh generation and dynamic manipulation is an integral part of many important scientific computing applications and must take place on the same platform with the rest of the application. Mesh generation is a memory-intensive application with major impact on the overall performance of the end-to-end field simulations-an important class of applications that could benefit greatly from a grid platform. Traditional approaches for generating, partitioning, and placing very large meshes have tow weakness when used on grids: (i) I/O and data movement due to mesh re-partitioning, for adaptive applications, is prohibitively expensive, and (ii) there is a trade-off between the performance of field solvers and the quality of the resulting elements and partitions.