MRI: Instrument Acquisition: Acquisition of Affordable Shared-memory Computing and Scalable Storage for Scientists and Engineers

MRI：仪器获取：为科学家和工程师获取经济实惠的共享内存计算和可扩展存储

• 批准号: 0619875
• 负责人: David Swanson
• 金额: $ 30万
• 依托单位: University of Nebraska-Lincoln
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2006
• 资助国家: 美国
• 起止时间: 2006-09-01 至 2009-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0619875&HistoricalAwards=false
• 关键词: MRI; Instrument; Acquisition; Affordable; Shared

This project, acquiring a cluster of multiprocessor nodes, aims at finding an affordable solution to the institution's oversubscribe power and storage for highly data intensive research. A shared memory resource is crucial for applications that do not scale well on a cluster, either due to the software paradigm used, or the large amount of contiguous memory address space required. The existing facility already has many users which take advantage of the available equipment to perform simulation and related experiments for research in Chemistry, Astronomy, Parallel Computing, Engineering Mechanics, Geosciences, etc. Hence, the acquisition contributes to fill two major gaps in the current UNL Research Computing Facility resources, shared memory resources and scalable storage, contiguous files system. The new infrastructure will help alleviate the bottleneck in the current systems.

这个项目，收购多处理器节点的集群，旨在找到一个负担得起的解决方案，该机构的超额认购的权力和存储高度数据密集型的研究。共享内存资源对于在集群上不能很好地扩展的应用程序至关重要，这可能是由于所使用的软件范例，也可能是由于需要大量的连续内存地址空间。现有的设施已经有许多用户利用现有的设备进行模拟和相关的实验研究化学，天文学，并行计算，工程力学，地球科学等，因此，收购有助于填补目前UNL研究计算设施资源，共享内存资源和可扩展存储，连续文件系统的两个主要差距。新的基础设施将有助于缓解现有系统的瓶颈。