MRI-R2: Acquisition of Shared Cluster and Database Computing Facilities at Wesleyan University

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。提案#：09-59856PI(S)：Starr，Francis W.；Beveridge，David，L.；Weir，Michael，P.机构：卫斯理大学标题：MRI-R2/ACQ：卫斯理大学共享集群和数据库计算设施项目建议：该项目，获得共享的、最先进的中型超级计算集群和数据库服务器，支持该机构密集的研究计算和相应的倡议。目前的中央设施不再能够满足密集的研究计算和教育方面的相应举措，悬而未决的工作数量在两年内增加了五倍就是明证。所请求的集群大约使中央超级计算设施的规模增加一倍，并为严重依赖关系数据库挖掘的基于信息学的研究增加了新的支持。天体物理、生物材料、液态化学物理、纳米技术、量子化学、分子生物物理、神经信息学和结构生物信息学的研究项目都需要高端计算资源，并越来越依赖高性能计算设施。所有这些领域都提供了涉及动手计算的课程。目前，大量教职员工和学生是活跃的研究用户，其中大约三分之二是学生；此外，每年有40-50名学生利用设施进行与课程相关的计算活动。根据与当前系统的主要用户和系统管理员的协商，新提议的集群由40-50个CPU节点组成，每个节点配备两个Intel Xeon四核处理器，包含8-16 GB的存储空间。InfiniBand互连将连接处理器，以促进并行应用。通过30-40 TB的RAID阵列为用户计算提供数据存储。专用数据库服务器具有双4核或6核处理器、大内存空间和用于存储数据库的高性能本地磁盘。该系统位于卫斯理科学大厦信息技术服务(ITS)楼层的翻新设施中，由其指定的系统管理员维护，并可通过当前的WESNET网络用于研究和教学需求。教职员工计算机咨询委员会提供学术监督，管理分配请求，并与其协调。该机构继续支持新系统的管理。更广泛的影响：该仪器既为教师研究服务，也为学生教育服务。它能够在不同的研究项目过程中产生新的科学，并影响大学正在进行的教学，作为一种学习工具，通过现有课程和参与教员领导的研究来发展学生的科学计算能力。为了进一步扩大教育和培训目标，在大学的资助下建立了一个新的科学计算和信息学中心，目的是促进所有教职员工和学生有效地使用新的高性能计算设施，并支持利用计算资源的课程的教育举措。SCIC还开展了一个暑期项目，支持本科生的研究和开发教学模块(与相关教师合作)，这些模块直接影响到整个课程的许多课程。集群和数据库设施也是本科生在计算密集型课程中的焦点和资源，包括最近批准的跨学科“信息学和建模证书计划”。这一跨学科项目有效地发挥了与主要领域并列的辅修领域的作用。该计划有两个要素，一个是计算方面的，一个是信息学方面的。

"This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5)." Proposal #: 09-59856PI(s): Starr, Francis W.; Beveridge, David, L.; Weir, Michael, P.Institution: Wesleyan University Title: MRI-R2/Acq: Shared Cluster and Database Computing Facilities at Wesleyan UniversityProject Proposed:This project, acquiring a shared, state-of-the-art, mid-sized super-computing cluster and database server, supports intensive research computing and corresponding initiatives at the institution. The current central facilities are no longer able to meet the intensive research computing and corresponding initiatives in education, as evidenced by a five-fold increase in the number of pending jobs over a period of two years. The requested cluster roughly doubles the size of the central super-computing facility, as well as adds new support for informatics-based research that relies heavily on relational database mining.Research programs in astrophysics, biomaterials, liquid state chemical physics, nanotechnology, quantum chemistry, molecular biophysics, neuroinformatics, and structural bioinformatics, all require high-end computing resources and have become increasingly dependent on high-performance computing facilities. All these areas offer courses that involve hands-on computation. Currently, a large number of faculty, staff and students are active research users, and roughly 2/3 of those are students; additionally, 40-50 students per year make use of facilities for course related computational activities.Based on consultations with the current primary users of the system and the systems administrator, the new proposed cluster consists of 40-50 CPU nodes, each with two Intel Xeon quad core processors, and containing 8-16 GB of storage. An infiniband interconnect will connect the processors to facilitate parallel applications. Data storage is provided for the user calculations via a RAID array of 30-40 TB. The dedicated database server has dual 4 or 6-core processors, a large memory footprint, and high performance local disks to store the database.The system is housed in the renovated facility on the Information Technology Services (ITS) floor of the Wesleyan Science Tower, maintained by a designated ITS systems manager and made available for research and instructional needs over the current WESNET network. A faculty Computer Advisory Committee provides academic oversight, manages allocation requests, and coordinates with ITS. The institution continues to support the administration of the new system.Broader Impacts: The instrument services both the faculty research and student education. It enables new science to be produced in the course of diverse research projects and impacts the ongoing instruction in the University, serving as a learning tool to develop student scientific computing proficiency both through existing courses and participating in faculty-led research. To further extend the educational and training goals, a new Scientific Computing and Informatics Center (SCIC) was developed with funding support from the University, aimed at facilitating the effective use of the new high-performance computing facility by all faculty and students, and supporting the educational initiatives of courses utilizing computational resources. The SCIC also operates a summer program supporting undergraduate research and developing teaching modules (in collaboration with relevant faculty) that directly impacts many courses across the curriculum. The cluster and database facility also serve as a focal point and resource for undergraduate students in computationally intensive courses of study, including a recently approved interdisciplinary 'Certificate Program in Informatics and Modeling'. This interdisciplinary program functions effectively as a minor field of study alongside major fields. The program has two elements, a computational and an informatics foci.