MRI: Acquisition of a GPU-Accelerated High Performance Computing and Visualization Cluster

MRI：获取 GPU 加速的高性能计算和可视化集群

• 批准号: 1229709
• 负责人: Inanc Senocak
• 金额: $ 55.54万
• 依托单位: Boise State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2012
• 资助国家: 美国
• 起止时间: 2012-10-01 至 2015-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1229709&HistoricalAwards=false
• 关键词: MRI; Acquisition; GPU; Accelerated; Performance

Science and engineering researchers across many disciplines have access to increasing amounts of computing power and research data available to support their work. State of the art cyberinfrastructure resources are needed to tackle large computational problems and make massive data accessible to promote superior analysis. This proposal plans for a 32- node GPU/CPU cluster with a storage array and a 5×8 foot tiled display to be located in a visualization theater setting. The equipment will support parallel computing and rendering, data storage, and high resolution imaging as researchers complete externally funded work. In addition, the instrument will be housed at a facility open to both university researchers as well as to local technology companies and partners, and provide partners with cyberinfrastructure access as well. This instrument will support research in multi-scale wind energy forecasting, data-driven modeling for threat reduction in chemical and biological defense, material characterization and modeling, snow hydrology and remote sensing, and fundamental studies on understanding the mechanisms in skeleton development in living systems and how living systems maintain complex three dimensional shapes. Boise State University is recognized as a CUDA Research Center for embracing GPU computing use across multiple research fields. The center?s goal is to develop and apply advanced numerical methods and computational algorithms to science and engineering, where rapid and real-time computations can transform current practice. This instrument would substantially increase computing power and play a key role in advancing research projects.The instrument will directly support Boise State's strategic goal to emerge as a metropolitan research university of distinction. exposure to cutting edge cyberinfrastructure and research can not only benefit researchers, but may also have profound effects on regional students who might otherwise never gain exposure. Faculty will also use the instrument to support externally funded education activities.

许多学科的科学和工程研究人员可以获得越来越多的计算能力和研究数据，以支持他们的工作。需要最先进的网络基础设施资源来解决大型计算问题，并使大量数据可供访问，以促进卓越的分析。该提案计划在可视化影院环境中部署一个带有存储阵列和5×8英尺平铺显示器的32节点GPU/CPU集群。随着研究人员完成外部资助的工作，该设备将支持并行计算和渲染、数据存储和高分辨率成像。此外，该仪器将放置在一个对大学研究人员以及当地技术公司和合作伙伴开放的设施中，并为合作伙伴提供网络基础设施接入。该仪器将支持以下领域的研究：多尺度风能预测、减少化学和生物防御威胁的数据驱动建模、材料特性和建模、雪水文学和遥感，以及了解生命系统骨架发育机制和生命系统如何保持复杂三维形状的基础研究。博伊西州立大学是公认的CUDA研究中心，在多个研究领域采用GPU计算。S的中心目标是开发先进的数值方法和计算算法，并将其应用到科学和工程中，在这些领域，快速和实时的计算可以改变目前的做法。该仪器将大大提高计算能力，并在推进研究项目中发挥关键作用。该仪器将直接支持博伊西州立大学成为一所著名的大都市研究型大学的战略目标。接触尖端的网络基础设施和研究不仅可以使研究人员受益，还可能对本地区的学生产生深远的影响，否则他们可能永远不会获得接触到的机会。教职员工还将使用该工具支持外部资助的教育活动。