MRI: Acquisition of a Supercomputing Cluster for Computational and Data-Intensive Applications in Science and Engineering

提案#：CNS07-22625PI(S)：Apon，Amy W.Bellaiche，Laurent；傅华祥；Pulay，Peter；Thompson。克雷格·W·机构：阿肯色大学费耶特斯维尔分校，AR 72701-2012标题：核磁共振/Acq.：用于计算和数据密集型应用的超级计算机集群科学与工程项目建议：该项目收购一个用于计算和数据密集型应用的集群，旨在支持计算机科学、物理和化学领域的项目，包括。电网容量规划，重点研究电网工作负荷和系统特性，建立仿真模型或性能分析、容量规划、调度评估和数据放置；设计低损耗、可调谐铁电材料，分辨有限频率和温度下的辩证损耗，以便微观理解损耗机理(S)；发现纳米结构材料的新物理，了解和发现新型纳米材料的新物理；开发并行计算方法以改进傅里叶变换库仑(FTC)和高效地实现耦合簇理论中的三重替换；数据索引和中间件，包括RFID中间件、合成数据生成服务、工作流网格子集、VMLab和技术过渡；用于纳米结构形成的自组装系统的计算设计用于喷雾冷却建模和纳米流体的高性能计算；纳米结构材料中界面行为的原子计算。计算集群是分布式存储并行应用程序的集成超级计算平台，也支持高性能和网格计算的协作课程。广泛的影响：基础设施有助于扩展活动，如研讨会和学术课程，吸引新用户，在该地区建立计算专业知识，以及培训学生。

Proposal #: CNS 07-22625PI(s): Apon, Amy W. Bellaiche, Laurent; Fu, Huaxiang; Pulay, Peter; Thompson. Craig W.Institution: University of Arkansas Fayettesville, AR 72701-1201Title: MRI/Acq.: Supercomputer Cluster for Computational & Data-Intensive Applications in Science & EngineeringProject Proposed:This project, acquiring a cluster for computational and data intensive applications, aims to enable projects in computer science, physics and chemistry, including. Grid Capacity Planning, focusing on grid workload and systems characterization and building simulation models or performance analysis, capacity planning, evaluation of scheduling, and data placement;. Approach to Design Low Loss, Tunable Ferroelectric Material, resolving dialectric loss at finite frequency and temperature for microscopic understanding of loss mechanism(s);. Discovering New Physics of Nanostructured Materials to understand and discover new physics in novel types of nanomaterials;. Developing Methods for Parallel Computing to Improve Fourier Transform Coulomb (FTC) and Efficient Implementation of Triple Substitutions in Coupled Cluster Theory;. Data Indexing and Middleware, including RFID middleware, Synthetic data generation service, Subsetting the workflow grid, VMlab, and Technology transition;. Computational Design of Self-assembly Systems for Nanostructured Formation;. High Performance Computing for Spray Cooling Modeling and Nanofluids; and. Atomistic Calculations of Interface Behavior in Nanostructured Materials.The computing cluster, an integrated supercomputing platform for distributed memory parallel applications, also enables collaborative courses in high-performance and grid computing.Broader Impact: The infrastructure contributes to outreach activities such as seminars and academic courses, attracting new users and building computational expertise in the region, as well as in the training of students.