II-EN: City University of New York - Computing Research Infrastructure

II-EN：纽约城市大学 - 计算研究基础设施

• 批准号: 0855217
• 负责人: Michael Kress
• 金额: $ 45.24万
• 依托单位: CUNY College of Staten Island
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-08-01 至 2012-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0855217&HistoricalAwards=false
• 关键词: II; EN; City; University; New

City University of New York (CUNY) researchers use high performance computing (HPC) to study air pollution in urban environments, develop techniques for visualization of scientific data, develop Monte Carlo-based algorithms for modeling of complex physical and transportation systems, and, with collaborators at the New York State Institute for Basic Research in Behavioral Disabilities (IBR), develop computational methods for autism studies. The investigators use a unique100-teraflops parallel heterogeneous HPC system, funded by NSF, which consists of 96 Intel 5570 quad-core processors connected to 96 NVidia Tesla graphics processing units. The Heterogeneous High Performance Computing System (H-HPCS), is tailor made for computations that are numerically intensive, vectorizable and highly parallel. CUNY scientists use H-HPCS to process atmospheric data obtained from satellites, lidar, and ground weather stations to assess pollution concentration and dispersion. Energy production and vehicular traffic within the City of New York generate large amounts of chemical and fine particulate pollution, which can induce or have adverse affects on residents, particularly those with cardio-respiratory illnesses. The H-HPCS techniques are developed to predict the dispersion of fixed point and mobile releases of chemicals and biological agents. IBR researchers conduct studies on the detection of autism related behavioral disorders using H-HPCS analysis of genomic data and video data obtained from motion capture systems. IBR and CUNY researchers collaborate on the development of these analysis techniques, which also have corollary applicability in the creative arts. The highly parallel nature of CUNY researcher?s models enable the use of the H-HPCS to study physical systems including mixing and transport of biological and physical properties by oceanographic flows, supersolids, atomic optical lattices, quantum phase transitions, and urban transportation systems.

纽约城市大学（CUNY）的研究人员使用高性能计算（HPC）来研究城市环境中的空气污染，开发科学数据可视化技术，开发基于蒙特卡罗的复杂物理和交通系统建模算法，并与纽约州行为残疾基础研究所（IBR）的合作者一起开发自闭症研究的计算方法。研究人员使用了一个独特的100万亿次/秒并行异构HPC系统，该系统由96个Intel 5570四核处理器和96个NVidia Tesla图形处理单元组成。异构高性能计算系统（H-HPCS）是为数字密集型、可向量化和高度并行的计算而量身定制的。纽约市立大学的科学家使用H-HPCS处理从卫星、激光雷达和地面气象站获得的大气数据，以评估污染浓度和扩散。纽约市的能源生产和车辆交通产生大量的化学物质和细颗粒物污染，这可能导致或对居民产生不利影响，特别是那些患有心肺疾病的人。H-HPCS技术用于预测化学制剂和生物制剂的定点和流动释放的分散。IBR研究人员利用H-HPCS分析基因组数据和从动作捕捉系统获得的视频数据，开展自闭症相关行为障碍检测的研究。IBR和纽约市立大学的研究人员合作开发这些分析技术，这些技术在创造性艺术中也有必然的适用性。纽约市立大学研究者的高度并行性？s的模型使H-HPCS能够用于研究物理系统，包括海洋流、超固体、原子光学晶格、量子相变和城市交通系统的生物和物理特性的混合和传输。