CRI: A Hierarchical Data Storage System for Large Data Simulation, Comparison, and Visualization

CRI：用于大数据模拟、比较和可视化的分层数据存储系统

• 批准号: 0551724
• 负责人: Charles Hansen
• 金额: --
• 依托单位: University of Utah
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2006
• 资助国家: 美国
• 起止时间: 2006-03-15 至 2011-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0551724&HistoricalAwards=false
• 关键词: CRI; Hierarchical; Data; Storage; System

This project, acquiring a large-scale hierarchical storage management (HSM) system, provides access to the increasingly large datasets produced and analyzed by the members and collaborators of the Scientific Computing and Imaging (SCI) Institute at Utah. This institute creates new techniques, tools, and systems by which scientists may solve computational and imaging problems in many disciplines. The expertise utilized in addressing the problems include image and signal processing, modeling, simulation, and visualization. The work exploits the multidisciplinary mission of SCI. A large-scale HSM solution provides the ability to generate canonical datasets, accessible by researchers within SCI and many collaborators outside. The field of scientific computing is constrained by the problems posed by dataset management and storage. Dataset size and complexity has grown in parallel with available compute resource. Centralized, ready access to the modeling, simulation, and analysis data is fundamental to success. Dataset complexity grows as a product of compute resource and algorithm efficiency; while research in complex problems relies on this ever-increasing efficiency for its progress. Although computational speeds have increased dramatically, the complexity of the models produced has also increased, so that while models and simulations have improved their research effectiveness, the actual computation time needed to generate a given simulation has not declined. An advanced storage architecture helps in streamlining the workflow for a variety of data intensive research projects.Broader Impact: The facility has a solid history of exporting its research via freely accessible datasets and software. Furthermore, SCI services many graduate and undergraduate students.

这个项目，收购一个大规模的分层存储管理（HSM）系统，提供了越来越大的数据集的访问和分析的成员和合作者的科学计算和成像（SCI）研究所在犹他州。该研究所创造了新的技术、工具和系统，科学家可以通过这些技术、工具和系统解决许多学科中的计算和成像问题。在解决这些问题时使用的专业知识包括图像和信号处理、建模、仿真和可视化。这项工作利用了SCI的多学科使命。大规模的HSM解决方案提供了生成规范数据集的能力，SCI内部的研究人员和外部的许多合作者都可以访问这些数据集。科学计算领域受到数据集管理和存储问题的限制。数据集的大小和复杂性与可用的计算资源并行增长。对建模、仿真和分析数据的集中、便捷的访问是成功的基础。数据集的复杂性随着计算资源和算法效率的增加而增加;而复杂问题的研究依赖于这种不断提高的效率。虽然计算速度显著提高，但所产生的模型的复杂性也增加了，因此，虽然模型和模拟提高了研究效率，但生成给定模拟所需的实际计算时间并没有减少。先进的存储架构有助于简化各种数据密集型研究项目的工作流程。更广泛的影响：该设施在通过免费访问的数据集和软件导出其研究方面有着坚实的历史。此外，SCI服务于许多研究生和本科生。