MRI: Acquisition of an Itanium Cluster for Grid Computing

EIA-0320895 Henry J. Neeman; Kelvin K. Droegemeier; Kyran D.放大图片作者：Dimitrios V. Papavassiliou;帕特里克L. SkubicUniversity of OklahomaMRI：收购用于网格计算的Itanium集群这一提案来自EPCoR州，为基于Itanium的网格寻求资金，旨在将各种计算领域的设施和研究集成到TeraGrid和预期的网络基础设施中。正在进行的气象学、高能物理学、地震、化学和航空航天工程项目将使用该资源来设计、开发和调试支持网格的软件。其中包括：大气发现建模环境（MEAD）大气发现关联环境（LEAD）数据网格工具的开发地震工程模拟网络（NEES）计算服务器开发基于网格的问题解决环境，用于油气藏模拟中通过多孔介质的多尺度多相流，和基于网格的飞机初步设计和优化设计环境。MEAD合作开发和调整网络基础设施，以实现仿真，数据挖掘/机器学习，和可视化的飓风和风暴使用TeraGrid。LEAD项目为真实的天气预报和相关研究开发网格资源。高能物理数据网格工具项目为科学分析社区开发动态网格工具，为他们的项目创建动态网格，并为RT交互式共享网格开发互补的合作实验室工具。NEESgrid项目将全国各地的地震研究人员与领先的计算资源联系起来，允许地理分布的协作。HiMUST项目为多孔介质的多尺度、多相、多物理场流动开发了一个问题解决环境。Grid-PRADO项目通过为设计工程师提供设计参数空间的大规模并行搜索，提供支持网格的飞机设计。该项目涉及两个少数民族机构，兰斯顿和霍华德大学的参与。

EIA-0320895Henry J. Neeman; Kelvin K. Droegemeier; Kyran D. Mish; Dimitrios V. Papavassiliou; Patrick L. SkubicUniversity of OklahomaMRI: Acquisition of an Itanium Cluster for Grid ComputingThis proposal from an EPCoR state, seeking funding for an Itanium based grid, aims at integrating the facility and the research in various computational fields into TeraGrid and contemplated Cyber-infrastructure. On-going projects in meteorology, high-energy physics, earthquake, chemical and aerospace engineering, will use the resource for design, development, and debugging Grid-enabled software. These include: Modeling Environment for Atmospheric Discovery (MEAD) Linked Environments for Atmospheric Discovery (LEAD) Development of Data Grid Tools Network for Earthquake Engineering Simulation (NEES) Computational Server Development of a Grid-based Problem Solving Environment for Multiscale, Multiphase Flow through Porous Media in Hydrocarbon Reservoir Simulation, and Grid-based Design Environment for Preliminary Aircraft Design and Optimization.The MEAD collaboration develops and adapts cyber-infrastructure to enable simulation, data mining/ machine learning, and visualization of hurricanes and storms using TeraGrid. The LEAD project develops Grid-enabled resources for real time (RT) weather forecasting and related research. The High Energy Physics Data Grid Tools project develops tools for dynamic workspaces for scientific analysis communities creating dynamic workspaces for their projects and complementary collaboratory tools for RT interactive sharing of workspaces. The NEESgrid project links earthquake researchers across the nation with leading-edge computing resources, allowing geographically distributed collaboration. The HiMUST project develops a problem solving environment for multiscale, multiphase, multiphysics flow of porous media. The Grid-PRADO project provides Grid-enabled aircraft design through massively parallel searching of design parameter spaces for design engineers.The project involves participation of two minority institutions, Langston and Howard Universities.