MRI: Acquisition of a Flexible High-Performance Computing System for Data and Compute Driven Scientific Discovery

MRI：获取灵活的高性能计算系统，用于数据和计算驱动的科学发现

• 批准号: 1625061
• 负责人: Michael Klein
• 金额: $ 50万
• 依托单位: Temple University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-10-01 至 2019-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1625061&HistoricalAwards=false
• 关键词: MRI; Acquisition; Flexible; Performance; Computing

This project, acquiring High-Performance Computing (HPC) resources for data and compute-driven scientific discovery, aims to replace and extend outdated current equipment at the core of the institution's HPC infrastructure. Since the proposed configuration is designed to particularly address deficiencies of the current infrastructure and thus adapt it to the needs of its widening user base, acquiring this instrumentation should enhance current HPC facilities and enable transformational research over a broad front. Central, shared, on-campus HPC facilities reduce barriers to access computational resources, especially for junior scientists, students, and faculty from disciplines that are not traditional HPC users. This equipment complements existing and future investments and helps drive related faculty and undergraduate research and educational activities across the College of Science and Technology (CST) and the entire Temple community. With an enrollment around 38,000 graduate and undergraduate students, the institution boasts of a very diverse student body. With its large shared memory pool, the requested machine will be the first of its kind in the Greater Philadelphia region that has a high concentration of high-tech and IT-related industries. They will obtain 1. A large shared memory machine with 9TB addressable RAM, 2. An extensible, scalable parallel storage appliance with 1.5PB capacity, and 3 A 60-node conventional Linux cluster tuned for small to medium size parallel calculations. Computational science, materials science, biophysics, data analytics, and genomics are major themes running through the most active areas of CST research. Various centers and institutes reflect these themes, such as the 1. Institute for Genomics and Evolutionary Medicine, 2, Institute for Computational Molecular Science, 3.Center for Data Analytics and Biomedical Informatics, 4. Center for Biophysics and Computational Biology, and 5. Center for Computational Design of Functional Layered Materials. CST's, in tandem with the Colleges of Engineering and Liberal Arts, the Schools of Business and Pharmacy, the Fox Chase Cancer Center, and the School of Medicine aim to broadly impact research in big data, nano-scale science, imaging, neuroscience, as well as energy and health.

该项目获取用于数据和计算驱动的科学发现的高性能计算 (HPC) 资源，旨在替换和扩展该机构 HPC 基础设施核心的过时的当前设备。由于拟议的配置旨在专门解决当前基础设施的缺陷，从而使其适应不断扩大的用户群的需求，因此获取该仪器应能增强当前的 HPC 设施，并在广泛的领域实现转型研究。中央共享的校园内 HPC 设施减少了获取计算资源的障碍，特别是对于非传统 HPC 用户学科的初级科学家、学生和教师而言。该设备补充了现有和未来的投资，并有助于推动科学技术学院 (CST) 和整个天普社区的相关教职员工和本科生研究和教育活动。该机构拥有约 38,000 名研究生和本科生，拥有非常多元化的学生群体。凭借其大型共享内存池，所请求的机器将是高科技和 IT 相关产业高度集中的大费城地区的第一台此类机器。他们将获得 1. 具有 9TB 可寻址 RAM 的大型共享内存机器，2. 具有 1.5PB 容量的可扩展、可扩展的并行存储设备，以及 3 专为中小型并行计算而调整的 60 节点传统 Linux 集群。计算科学、材料科学、生物物理学、数据分析和基因组学是 CST 研究最活跃领域的主要主题。各个中心和研究所都反映了这些主题，例如 1. 基因组学和进化医学研究所，2. 计算分子科学研究所，3. 数据分析和生物医学信息学中心，4. 生物物理学和计算生物学中心，5.功能层状材料计算设计中心。 CST 与工程学院和文学院、商学院和药学院、福克斯蔡斯癌症中心和医学院合作，旨在广泛影响大数据、纳米科学、成像、神经科学等领域的研究，以及能量和健康。

项目成果

Discovering research articles containing evolutionary timetrees by machine learning

通过机器学习发现包含进化时间树的研究文章

• DOI: 10.1093/bioinformatics/btad035
• 发表时间: 2023-01-01
• 期刊: Bioinformatics
• 影响因子: 5.8
• 作者: Marija Stanojevic;Jovan Andjelkovic;Adrienne E Kasprowicz;Louise A. Huuki;Jennifer Chao;S. Hedges;Sudhir Kumar;Z. Obradovic
• 通讯作者: Z. Obradovic