MRI: Acquisition of Computationally Intensive Research in High Energy Physics

MRI：获取高能物理计算密集型研究

• 批准号: 0723103
• 负责人: Jeffrey Nelson
• 金额: $ 25.43万
• 依托单位: College of William and Mary
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-08-15 至 2009-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0723103&HistoricalAwards=false
• 关键词: MRI; Acquisition; Computationally; Intensive; Research

High energy physics has a long and close association with high performance scientific computation. The data samples collected with massive particle detectors of experimental HEP require significant data handling and processing loads. The push for grid-based computing was significantly driven by the phenomenal computing needs expected for the LHC experiments that will soon be coming online at CERN in Geneva, Switzerland. Even categories of experiments that traditionally do not demand large computation resources, such as neutrino experiments, are now producing high statistics datasets that are requiring significant computational resources for analysis and interpretation of the results of recent experiments.On the theoretical side, the physics of the strong interaction (QCD) had been analytically intractable for a wide range of problems of current interest. Lattice QCD is a tool that allows us to investigate the dynamics of QCD, understand how certain low energy phenomena arise, and compute the consequences of the Standard Model. It is expected to play an important role in supporting the experimental effort and in discovering new physics. The computational demands for the high precision, phenomenologically interesting lattice QCD calculations are such that high performance massively parallel supercomputers are required. However, small scale resources are also critical for developing new ideas and training graduate and undergraduate students. In addition, small scale resources with enough local storage can play an important role in the analysis of the gauge field configurations and quark propagators generated on the big machines. This part of the analysis is usually done by small groups located at universities.This proposal addresses the computational needs of the W&M neutrino group and the larger neutrino community for simulations of neutrino interactions in the MINOS neutrino detectors currently operation at Fermilab in Batavia, Illinois and the Soudan Underground Lab in northern Minnesota. Concurrently, this proposal will provide a facility for performing small pilot projects in lattice QCD and for supporting the data analysis requirements of the large computations performed on the major Lattice QCD machines. In effect, this proposal enables the PIs and W&M students to play a significant role in the larger Lattice QCD community and allow them access to the results of the major computations in this area.

高能物理与高性能科学计算有着长期而密切的联系。 实验高能物理的大质量粒子探测器采集的数据样本需要大量的数据处理和处理负载。 对网格计算的推动主要是由即将在瑞士日内瓦的欧洲核子研究中心上线的大型强子对撞机实验所预期的惊人的计算需求所推动的。 即使是那些传统上不需要大量计算资源的实验，如中微子实验，现在也产生了高统计数据集，这些数据集需要大量的计算资源来分析和解释最近的实验结果。在理论方面，强相互作用（QCD）物理学在分析当前感兴趣的广泛问题时一直是难以解决的。 格点QCD是一种工具，它使我们能够研究QCD的动力学，理解某些低能现象是如何产生的，并计算标准模型的结果。 预计它将在支持实验努力和发现新物理方面发挥重要作用。 高精度的计算需求，唯象有趣的晶格QCD计算是这样的，高性能的大规模并行超级计算机是必需的。 然而，小规模的资源也是开发新思想和培养研究生和本科生的关键。 此外，具有足够局域存储的小尺度资源可以在分析大机器上产生的规范场组态和夸克传播子中发挥重要作用。 这部分的分析通常是由位于university.This建议解决的计算需求的W M中微子组和更大的中微子社区的中微子相互作用的模拟在MINOS中微子探测器目前在费米实验室在巴达维亚，伊利诺伊州和Soudan地下实验室在北方明尼苏达州。 同时，这项建议将提供一个设施，用于执行小型试点项目在格QCD和支持大型计算的数据分析要求上进行的主要格QCD机。 实际上，这一建议使PI和W M的学生在更大的晶格QCD社区中发挥重要作用，并允许他们访问该领域的主要计算结果。