Loading...
Please wait, while we are loading the content...
Similar Documents
Integrating Large-scale Distributed and Parallel HPC ( DPHPC ) Applications Using a Component-Based Architecture
| Content Provider | Semantic Scholar |
|---|---|
| Author | Wang, Nanbor Hamill, Paul Liu, Fang |
| Copyright Year | 2008 |
| Abstract | Component-based software engineering (CBSE) is now a widel y accepted paradigm for developing large-scale commercial s oftware. The Common Component Architecture (CCA) and its associated Babel tool suite are designed to enable CBSE for High Performance Computing (HPC) scientific applications. Many scient ific applications have adopted the CBSE paradigm and demonstrat ed its effectiveness using CCA. Integrating multiple distributed HPC applications into a l argescale system allows the overall system to take advantage of s cattered resources and to provide alternative strategies in en hancing system’s overall parallelism and performance. CCA is wellsuited to facilitate such applications as it supports both paralle l nd distributed interaction models. For example, applications ca n utilize remote visualization and coordination of large-scale phys ics imulations. In this paper, we demonstrate how VORPAL, a particle-in-cel l (PIC) plasma simulation code, takes advantage of a Distribu ted and Parallel HPC (DPHPC) approach by constructing various data analysis strategies to perform simulation using distributed co mputing resources. We provide both qualitative and quantitative co mparisons of the capabilities (such as reusability, composabil ity, etc.) and speedup we achieved using a DPHPC approach with a compone ntbased framework. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://eprints.cca-forum.org/125/1/CBHPC-08.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |