MRI: Acquisition of a High Performance Computing Cluster to Support Multidisciplinary Big Data Analysis and Modeling

MRI：收购高性能计算集群以支持多学科大数据分析和建模

• 批准号: 1429518
• 负责人: Long-zhuang Li
• 金额: $ 39.9万
• 依托单位: Texas A&M University Corpus Christi
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-09-01 至 2017-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1429518&HistoricalAwards=false
• 关键词: MRI; Acquisition; Performance; Computing; Cluster

This project, acquiring a High Performance Computing (HPC) cluster--the Corpus Christi High Performance cluster (CCHP)--supports large-scale data analysis and modeling research and research training across a broad variety of science and engineering technology disciplines in coastal and environmental studies. The CCHP cluster enables research projects ranging from computer science, life science, geographical information systems (GIS), remote sensing, to atmospheric science. These include:- Higher order tensor decomposition for big data,- Population Genomics of non-model species,- Geospatial crowdsourcing for natural disaster response,- Large-scale analytics of airborne and satellite remote sensing data and UAS imagery,- Global weather and climate analysis.CCHP contains compute nodes, GPU (graphical processing unit) nodes, and shared network storage, connected through InfiniBand switches in a fat tree topology to support high bandwidth low latency data communication (critical for HPC applications) while providing massive parallel computation on graphics processor cores. Moreover, the on-board Gigabit Ethernet ports with switch connection can support large data sets transmission which enables research in real time simulation and modeling. CCHP enables exploring parallel processing to process real large data sets in the higher order tensor decomposition, which is a basis for many data mining tasks including clustering, trend detection, and anomaly detection. This cluster is a necessary computational tool for biologists to use statistics effectively to realize the promise and power of societally relevant hypotheses in massively parallel nucleotide sequencing. The GIS researchers can utilize the CCHP cluster to advance geospatial crowdsourcing solution in case of natural disaster by providing quicker and more accurate geometric information. The cluster also enables remote sensing scientists to identify the relationship between environmental conditions, including land cover and use and rates of freshwater inflow, and attack the problems caused by the low accuracy of acquired unmanned aerial systems (UAS) images for precision agriculture. Furthermore, processing much longer time series of the global model and satellite data record, atmospheric scientists will be able to easily expand their research projects from regional to global scale.The impact of the CCHP cluster will be felt in many domains, especially on coastal and environmental studies that impacts society in general, impacting weather and climate model reliability and prediction skills in ecology, evolution, fisheries, conservation, and genetics. Moreover, curriculum materials obtained from the projects serviced by the instrumentation will enhance education and student learning at all levels, including K-12. These will impact existing courses and contribute to the creation of new ones.

该项目收购高性能计算(HPC)集群--Corpus Christi高性能集群(CCHP)--支持在沿海和环境研究的各种科学和工程技术学科中进行大规模数据分析和建模研究以及研究培训。CCHP集群支持从计算机科学、生命科学、地理信息系统(地理信息系统)、遥感到大气科学的各种研究项目。其中包括：-大数据的高阶张量分解，-非模式物种的种群基因组学，-自然灾害应对的地理空间众包，-机载和卫星遥感数据以及UAS图像的大规模分析，-全球天气和气候分析。CCHP包含计算节点、GPU(图形处理单元)节点和共享网络存储，通过InfiniBand交换机以胖树拓扑结构连接，以支持高带宽低延迟数据通信(对HPC应用至关重要)，同时在图形处理器内核上提供大规模并行计算。此外，带交换机连接的车载千兆以太网端口可以支持大数据集传输，这使得实时仿真和建模研究成为可能。CCHP能够在高阶张量分解中探索并行处理来处理真实的大数据集，这是许多数据挖掘任务的基础，包括聚类、趋势检测和异常检测。对于生物学家来说，这个簇是一个必要的计算工具，可以有效地使用统计学来实现大规模并行核苷酸测序中与社会相关的假设的前景和力量。通过提供更快、更准确的几何信息，地理信息系统研究人员可以利用CCHP集群来推进自然灾害情况下的地理空间众包解决方案。该集群还使遥感科学家能够确定包括土地覆盖和使用在内的环境条件与淡水流入速度之间的关系，并解决因获取的用于精准农业的无人机系统(UAS)图像精度较低而造成的问题。此外，通过处理更长的全球模式时间序列和卫星数据记录，大气科学家将能够轻松地将他们的研究项目从区域扩展到全球范围。CCHP集群的影响将在许多领域感受到，特别是在影响整个社会的沿海和环境研究方面，影响到天气和气候模型的可靠性以及生态学、进化论、渔业、养护和遗传学方面的预测技能。此外，从仪器设备服务的项目中获得的课程材料将加强各级教育和学生的学习，包括K-12。这些将影响现有的课程，并有助于创建新的课程。