Showing papers by "Joseph E. Flaherty published in 2004"
TL;DR: By detecting discontinuities in such variables as density or entropy, limiting may be applied only in these regions; thereby, preserving a high order of accuracy in regions where solutions are smooth.
404 citations
20 Jun 2004
TL;DR: This work focuses on partitioning and dynamic load balancing on hierarchical procedures implemented within the Zoltan Toolkit, guided by DRUM, the Dynamic Resource Utilization Model, and shows that hierarchical partitionings are competitive with the best traditional methods on a small hierarchical cluster.
Abstract: Cluster and grid computing has made hierarchical and heterogeneous computing systems increasingly common as target environments for large-scale scientific computation. A cluster may consist of a network of multiprocessors. A grid computation may involve communication across slow interfaces. Modern supercomputers are often large clusters with hierarchical network structures. For maximum efficiency, software must adapt to the computing environment. We focus on partitioning and dynamic load balancing, in particular on hierarchical procedures implemented within the Zoltan Toolkit, guided by DRUM, the Dynamic Resource Utilization Model. Here, different balancing procedures are used in different parts of the domain. Preliminary results show that hierarchical partitionings are competitive with the best traditional methods on a small hierarchical cluster.
26 citations
Proceedings Article•
01 Jan 2004
TL;DR: SALSA-MPI as mentioned in this paper is an actor programming language approach to scientific computing that extends MPI with a checkpointing and migration API and a runtime system that manages both periodic checkpoints and process or application migration.
Abstract: Over the last two decades, efficient message passing libraries have been developed for parallel scientific computation. Concurrently, programming languages have been created supporting dynamically reconfigurable distributed systems over the heterogeneous Internet. In this paper, we introduce SALSA-MPI, an actor programming language approach to scientific computing that extends MPI with a checkpointing and migration API and a runtime system that manages both periodic checkpoints and process or application migration. The goal is to enable dynamic network reconfiguration and load balancing without sacrificing application performance or requiring extensive code modifications. As driving technology for this effort of unifying parallel and distributed computing, we plan to use adaptive solvers of partial differential equations. Fields as diverse as fluid dynamics, material science, biomechanics, and ecology make use of parallel adaptive computation, but target architectures have traditionally been supercomputers and tightly-coupled clusters, SALSA-MPI is intended to allow these computations to make efficient use of more distributed and dynamic computing resources.
8 citations
15 Aug 2004
TL;DR: Melt flows associated with a Czochralski crystal growth process was investigated to better understand the transition from a steady laminar regime to an unsteady one as the Grashof number increases, and suggested the Reynolds quasi-steady assumption is valid.
Abstract: Melt flows associated with a Czochralski crystal growth process was investigated to better understand the transition from a steady laminar regime to an unsteady one as the Grashof number increases. The kinetic energy of the flow as a function of time was examined as an indication of stability. Three-dimensional solutions were interpolated onto a two-dimensional unstructured mesh to compute the Reynolds average mean flow and its fluctuations. Our simulations showed that the transition to unsteady three-dimensional flow begins at a Grashof number of approximately 3.0 million. At higher Grashof numbers (e.g., 6.6 million), the melt flow is fully unsteady, three-dimensional turbulent flow. The simulations further indicated that the melt flow at a Grashof number of 6.6 million is statistically stable, which suggested the Reynolds quasi-steady assumption is valid in this case.