MRI: Acquisition of a Massive Database to Accelerate Data Science Discovery

该项目由主要研究仪器和刺激竞争性研究既定计划（EPSCoR）项目共同资助。该项目资助佛蒙特大学 (UVM) 建设 DataMountain，这是一个用于高性能计算的大型数据库集群。大内存机器将增强佛蒙特州高级计算核心，这是一个支持佛蒙特州 500 多名科学家研究的虚拟实验室。 随着许多领域从数据稀缺环境转变为数据丰富环境，许多重要的研究领域都将从这台新机器中受益，包括对成瘾、精神疾病、气候变化、药物发现、食品系统和在线错误信息传播的研究。 DataMountain 将允许快速访问巨大的数据集，支持多个需要计算能力和速度来有效分析、描述和解释快速增长的数据集的项目。DataMountain 将使 UVM 可用于计算研究的最大随机存取存储器机器增加近两个数量级，加速需要快速读写的大规模数据驱动研究，并促进目前在现有硬件下不可能实现的广泛而多样化的重要科学研究。它还将增强高性能计算集群 BlueMoon 和 DeepGreen 的功能，这两个集群分别致力于并行处理和机器学习。例如，该机器将允许通过 http://storywrangling.org 和 http://hedonometer.org 交互式访问超过 50 TB 的社交媒体数据，以便及时分析与 COVID-19 大流行相关的人口规模身心健康数据的变化。此外，DataMountain 将允许大幅提高计算化学模拟的空间和时间分辨率，以数据驱动设计下一代抗菌肽，以对抗抗生素耐药性。 DataMountain 还将在未来十年内探索与美国 10,000 名青少年相关的 PB 级功能磁共振成像、遗传、任务表现和调查数据。此外，该机器将加速利用无人机监视成像进行树冠评估的研究，促进全球作物农业多样性和营养结果的网络科学建模，并帮助量化COVID-19大流行对粮食不安全的影响。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project is jointly funded by the Major Research Instrumentation and the Established Program to Stimulate Competitive Research (EPSCoR) programs. The project funds construction of DataMountain, a massive database cluster for high performance computing at the University of Vermont (UVM). The large-memory machine will enhance the Vermont Advanced Computing Core, a virtual laboratory supporting the research of over 500 scientists in the state of Vermont. With so many fields transitioning from data-scarce to data-rich environments, many important research areas will benefit from this new machine including research into addiction, mental illness, climate change, drug discovery, food systems, and the spread of online misinformation. DataMountain will allow for fast access to enormous datasets, supporting several projects that require computational power and speed to effectively analyze, describe, and explain rapidly growing datasets.DataMountain will increase by nearly two orders of magnitude the largest random access memory machine available for computational research at UVM, accelerating large-scale data-driven research requiring rapid reading and writing, and facilitating a broad and diverse set of important scientific investigations not currently possible given the existing hardware. It will also enhance the functionality of the high performance computing clusters BlueMoon and DeepGreen, which are dedicated to parallel processing and machine learning respectively. For example, the machine will allow for interactive access to over 50 terabytes of social media data through http://storywrangling.org and http://hedonometer.org for timely analysis of changes related to the COVID-19 pandemic in population-scale physical and mental health data. In addition, DataMountain will allow for massive increases in the spatial and temporal resolution of computational chemistry simulations being performed for data-driven design of next-generation antimicrobial peptides to combat antibiotic resistance. DataMountain will also enable exploration of petabytes of fMRI, genetic, task performance, and survey data associated with 10,000 adolescents across the United States over the next decade. In addition, the machine will accelerate research using unmanned aerial surveillance imaging for tree canopy assessments, facilitate network science modeling of agricultural diversity of crops and nutritional outcomes globally, and help quantify the impacts of the COVID-19 pandemic on food insecurity.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.