US-Egypt Cooperative Research: RAMSys: Collaborative Metacomputing System

0138760 Greenshields描述：该奖项是为了支持一个合作项目，该项目由位于康涅狄格州斯托尔斯的康涅狄格大学计算机科学与工程系的Ian Greenshields博士和Reda Ammar博士以及埃及开罗电子研究所（ERI）主席Ayman El-Dessouki博士共同完成。他们建议进行研究，试图在开罗建立ERI，作为埃及网格计算的卓越中心，连接埃及和美国的计算资源。该项目将包括三个阶段：建立阶段，在ERI建立一个小型的美国-埃及集群，以及现有的网格中间件;教育阶段，他们将共同为埃及（以及美国）创建一个在线教育资源。在研究阶段，他们将与计算医学的一个合作项目一起使用小集群，以评估与时间差相关的负载特性，以及低带宽网络连接对问题粒度和计算时间与数据传输时间之比的影响。范围：将广泛分布的计算资源聚合到无缝的大规模并行虚拟机中（网格计算机）有潜力提供成本-高效的超级计算，并通过在大规模虚拟数据空间中共享数据来鼓励国际科学合作。然而，发展中国家的科学家往往无法使用大型网格计算机，也无法获得使他们能够利用廉价的商用工作站集群作为协作网格内的本地节点的那种信息技术支持。 拟议中的发展可以为埃及科学界提供一个中等规模的网格计算机，由于埃及和美国之间的时差，在美国，与埃及科学家合作的美国站点的机器可以在美国端的非工作时间由埃及利用，同样，美国在埃及端的非工作时间使用埃及机器也是如此。美国PI和埃及合作者在这一研究领域拥有丰富的经验。康涅狄格大学的一名研究生将在埃及从事该项目。 该提案符合INT的目标，即支持在共同科学感兴趣的领域进行合作研究。

0138760GreenshieldsDescription: This award is to support a collaborative project by Dr. Ian Greenshields and Dr. Reda Ammar, Department of Computer Science and Engineering, University of Connecticut, Storrs, Connecticut, and Dr. Ayman El-Dessouki, President of the Electronics Research Institute (ERI), Cairo, Egypt. They propose to do research, which seeks to establish the ERI in Cairo as a Center of Excellence in Egypt in grid computing, linking Egyptian and US computation resources. The project will include three phases: an establishment phase, in which a small US-Egyptian cluster is established at ERI, together with extant grid middleware; an educational phase, in which they will jointly create an on-line educational resource for Egyptian (and also for U.S.) scientists regarding grid computing; and a research phase, in which they will exercise the small clusters with a collaborative project in computational medicine to assess loading characteristics relative to time differences and the effect of low bandwidth network connections relative to problem granularity and the ratio of compute time to data transfer time.Scope: The aggregation of widely distributed computing resources into a seamless massively parallel virtual machine (grid computer) has the potential for providing cost-effective supercomputing and for encouraging international scientific collaboration by the sharing of data in a massive virtual data space. However, scientists in developing countries tend not to have access either to massive grid computers or to the kind of IT support which allows them to exploit inexpensive clusters of commodity workstations as local nodes within a collaborative grid. The proposed development can provide the Egyptian scientific community with a medium-scale Grid computer, and because of the time differences between Egypt and the U.S., machines at U.S. sites which collaborate with Egyptian scientists can be harnessed by Egypt during the off-hours at the U.S. end, and the same holds for U.S. use of Egyptian machines during off-hours at the Egyptian end. The US PI and the Egyptian collaborator have extensive experience in this area of research. One graduate student from the University of Connecticut will work on the project in Egypt. This proposal meets the INT objective of supporting collaborative research in areas of mutual scientific interest.