MRI: Acquisition of a High Performance Cluster for the University of Maine Scientific Grid Portal

• Main Authors: Dickens, Phillip M., Zhu, Yifeng, Segee, Bruce E., Fastook, James L., Chawathe, Sudarshan S.
• Format: Text
• Language: unknown
• Published: DigitalCommons@UMaine 2009
• Subjects: Research instrumentation; Computer Sciences; The Portal; Ice Sheet
• Online Access: https://digitalcommons.library.umaine.edu/orsp_reports/310; https://digitalcommons.library.umaine.edu/cgi/viewcontent.cgi?article=1317&context=orsp_reports

This project, acquiring a cluster to establish a scientific grid portal in Maine, aims to enable projects requiring large datasets. The work makes available to the wider community results such as widely-used whole-ice sheet models, tools for climate change research, prototype versions of object-based caching system (bundled with MPI-IO implementation developed at Argonne National Lab), the data management system, real-time animations, videos, etc. Additionally, the portal provides the larger community the compute power, storage capacity, and rendering engine to execute very high-resolution models, and receive animations and other visualized information in real time.Broader Impact: The infrastructure enhances understanding of global issues and contributes in the development of educational tools for K-12 students. The scientific grid portal contributes in the dissemination of important scientific discoveries. The portal also provides a show-case for research being performed in the state.