MRI: Acquisition of a Supercomputing Cluster for Data-Intensive Computational Research using the Open Science Grid (OSG) Cyberinfrastructure

MRI：使用开放科学网格 (OSG) 网络基础设施获取用于数据密集型计算研究的超级计算集群

• 批准号: 0922923
• 负责人: Akhtar Mahmood
• 金额: $ 20.03万
• 依托单位: Bellarmine University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2010
• 资助国家: 美国
• 起止时间: 2010-04-01 至 2011-03-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0922923&HistoricalAwards=false
• 关键词: MRI; Acquisition; Supercomputing; Cluster; Data

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). Bellarmine University will acquire a 64-node Supercomputing Cluster with approximately 22TB of RAID6 disk storage space to conduct grid-enabled computational research in High Energy Physics using the Open Science Grid (OSG) cyberinfrastructure. The collaborative research activities will be carried out in close partnership with University of Oklahoma, State University of New York at Albany, Purdue University at Calumet, Southwestern University, and a Historically Black Minority-Serving institution, Benedict College. Using the grid middleware resources, this partnership will essentially create a virtual organization that will link all the research participants by bringing together geographically and organizationally dispersed computational resources and people from the six institutions from six different states by developing a platform for shared systematic use of the supercomputing cluster and effective utilization of computational tools. We plan to search for the Higgs Boson and SUSY (Supersymmetric) particles in specific decay channels using the LHC (Large Hadron Collider) data. The physics studies will first be carried out and tested on detailed simulations of particles and their interactions with the detector using the GEANT monte-carlo package. This multi-institutional research team will focus on the development of evolutionary algorithms and computational code to carry out the complex data analyses tasks by studying the various detector parameters - focusing on data reduction and filtering methodologies and pattern recognition techniques for extracting ?discovery level signals? from petabyte-scale datasets. The availability of a supercomputing cluster will serve as a powerful stimulus for the recruitment of undergraduate students in the sciences, especially from the underrepresented groups. The computational research activities with OSG will not only have a significant and a broad societal impact on the educational training of undergraduate students of all races and ethnicity at Bellarmine University, but also has the potential to bring in a new era of grid-enabled scientific research capability at this institution via infusion of this new knowledge. The grid-enabled supercomputing cluster will create precisely the kind of exciting learning environment that will provide the next generation of students at all seven participating institutions with an accessible entry to research and education at the frontiers of computational science, information technology, and cluster/grid computing. These trained students will acquire a solid computational background with broad technical skills and expertise in high performance parallel cluster computing techniques and grid middleware technologies. Using this supercomputing facility, we plan to incorporate the grid-enabled research activities into the academic curriculum by developing an innovative interdisciplinary course in ?Computational Science using Grid Technology.? Additionally, we plan to host a hands-on OSG Summer Grid Workshop/School at Bellarmine University for the undergraduate students from the regional public, private, and minority-serving institutions in Kentucky.

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。贝拉明大学将获得一个64节点的超级计算集群，拥有大约22 TB的RAID 6磁盘存储空间，使用开放科学网格（OSG）网络基础设施在高能物理领域进行网格计算研究。 合作研究活动将与俄克拉荷马州大学、奥尔巴尼的纽约州立大学、卡卢梅的普渡大学、西南大学和历史上为黑人少数民族服务的机构本尼迪克特学院密切合作进行。使用网格中间件资源，这种伙伴关系将基本上创建一个虚拟组织，将所有的研究参与者通过将地理上和组织上分散的计算资源和来自六个不同州的六个机构的人员聚集在一起，通过开发一个平台来共享超级计算集群的系统使用和有效利用计算工具。我们计划利用大型强子对撞机（LHC）的数据在特定的衰变通道中寻找希格斯玻色子和超对称粒子。物理学研究将首先进行，并使用GEANT蒙特-卡罗软件包对粒子及其与探测器的相互作用进行详细模拟和测试。 这个多机构研究团队将专注于进化算法和计算代码的开发，通过研究各种探测器参数来执行复杂的数据分析任务-专注于数据简化和过滤方法以及用于提取的模式识别技术。发现级信号从PB级的数据集。 超级计算集群的可用性将成为招收理科本科生的强大刺激因素，特别是来自代表性不足的群体。与OSG的计算研究活动不仅将对贝拉明大学所有种族和民族的本科生的教育培训产生重大和广泛的社会影响，而且有可能通过注入这种新知识，在该机构带来网格支持的科学研究能力的新时代。支持网格的超级计算集群将精确地创建一种令人兴奋的学习环境，为所有七个参与机构的下一代学生提供一个可访问的入口，以研究和教育计算科学，信息技术和集群/网格计算的前沿。 这些受过培训的学生将获得坚实的计算背景，在高性能并行集群计算技术和网格中间件技术方面具有广泛的技术技能和专业知识。使用这个超级计算设施，我们计划将网格使能的研究活动纳入学术课程，通过开发一个创新的跨学科课程？使用网格技术的计算科学。此外，我们计划在贝拉明大学为来自肯塔基州地区公立、私立和少数民族服务机构的本科生举办一个动手操作的OSG夏季网格研讨会/学校。