MRI: Acquisition of Hybrid CPU/GPU High Performance Computing and Storage for STEM Research and Education at San Jose State University

MRI：圣何塞州立大学为 STEM 研究和教育采购混合 CPU/GPU 高性能计算和存储

• 批准号: 1626645
• 负责人: Sen Chiao
• 金额: $ 90.08万
• 依托单位: San Jose State University Foundation
• 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: 2016-08-01 至 2020-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1626645&HistoricalAwards=false
• 关键词: MRI; Acquisition; Hybrid; CPU; GPU

This award is to acquire a state-of-the-art high-performance computing (HPC) facility at San Jose State University (SJSU). The HPC system will provide faculty and students regular access to a modern, on-campus computing facility for computational science and engineering research. As a key hub for STEM fields in the San Francisco Bay Area, this facility will promote the progress of science and engineering, as well as offer a wide diversity of experiences for our students through required laboratory courses and research opportunities. This multidisciplinary and collaborative project involves faculty and students from biological science, chemistry, computer science, aerospace engineering, computer engineering, meteorology and climate science, physics, astronomy, mathematics, and statistics. The HPC will further enhance SJSU?s capability as a focal point in training members of the biotechnology/ pharmaceutical and information technology workforce in the Bay Area. This facility will also contribute to attracting students from underrepresented groups as well as local community colleges into STEM fields at SJSU. It is anticipated that more than 200 students from SJSU each year in STEM related courses and activities will benefit from such an instrument.Comprising a hybrid central processing unit (CPU)/graphics processing unit (GPU) HPC built using 1696 compute cores and a 1.0-petabyte High Performance Storage System (HPSS). This system will be used for computational analysis, data-intensive research, rich media, three-dimensional computer modeling, data mining, and large-scale simulation. The projects that are poised to commence include: on-demand numerical weather prediction, assimilation, and analysis (Atmospheric Science); dynamical modeling of orbits and dark matter in gas-poor galaxies (Physics and Astronomy); computational modeling of Tat peptide mutants binding to BIV TAR RNA and protein-protein interfaces (Biochemistry); quantum mechanical properties of materials in the atomic scale (Physics and Astronomy); guidance and trajectory optimization strategies in presence of wind, and spacecraft and orbital trajectory optimization (Aerospace Engineering); genomic assessment of adaptation, and pharmacological and evolutionary perspective on bioactive compounds in marine invertebrates (Biological Science); high-resolution simulations of weather phenomena, dust transport, and climate on Mars (Planetary Science); and efficient algorithms for modeling large amount of data in high dimensions (Mathematics and Statistics). The HPC infrastructure and associated user group will enable many further follow-up projects, including cross-disciplinary collaborations as well as participation from the wider SJSU community.

该奖项旨在收购圣何塞州立大学 (SJSU) 最先进的高性能计算 (HPC) 设施。 HPC 系统将为教师和学生提供定期访问现代化校园计算设施的机会，以进行计算科学和工程研究。作为旧金山湾区 STEM 领域的重要枢纽，该设施将促进科学和工程的进步，并通过必修的实验室课程和研究机会为我们的学生提供广泛多样的体验。这个多学科合作项目涉及生物科学、化学、计算机科学、航空航天工程、计算机工程、气象和气候科学、物理学、天文学、数学和统计学的教师和学生。 HPC 将进一步增强 SJSU 作为湾区生物技术/制药和信息技术人员培训中心的能力。该设施还将有助于吸引弱势群体以及当地社区学院的学生进入 SJSU 的 STEM 领域。预计每年有超过 200 名 SJSU 学生在 STEM 相关课程和活动中受益于这种仪器。该仪器包括一个混合中央处理单元 (CPU)/图形处理单元 (GPU) HPC，使用 1696 个计算核心和一个 1.0 构建-PB 高性能存储系统（HPSS）。该系统将用于计算分析、数据密集型研究、富媒体、三维计算机建模、数据挖掘和大规模模拟。即将启动的项目包括：按需数值天气预报、同化和分析（大气科学）；贫气体星系中轨道和暗物质的动力学模型（物理学和天文学）； Tat 肽突变体与 BIV TAR RNA 和蛋白质-蛋白质界面结合的计算模型（生物化学）；原子尺度材料的量子力学特性（物理学和天文学）；有风情况下的制导和轨迹优化策略，以及航天器和轨道轨迹优化（航空航天工程）；适应的基因组评估以及海洋无脊椎动物生物活性化合物的药理学和进化观点（生物科学）；火星上天气现象、尘埃传输和气候的高分辨率模拟（行星科学）；用于对高维大量数据进行建模的高效算法（数学和统计）。 HPC 基础设施和相关用户组将支持许多进一步的后续项目，包括跨学科合作以及更广泛的 SJSU 社区的参与。