Collaborative Research: Multiscale Software for Quantum Simulations in Nano Science and Technology

合作研究：纳米科学技术中量子模拟的多尺度软件

• 批准号: 0749320
• 负责人: Jerzy Bernholc
• 金额: $ 102万
• 依托单位: North Carolina State University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2007
• 资助国家: 美国
• 起止时间: 2007-10-01 至 2013-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0749320&HistoricalAwards=false
• 关键词: Collaborative; Research; Multiscale; Software; Quantum

TECHNICAL SUMMARY:This award is made on a proposal submitted to the PetaApps Solicitation. The Office of Cyberinfrastructure, the Computer and Information Science and Engineering Directorate, and the Divisions of Materials Research, Physics, and Chemistry and the Office of Multidisciplinary Activities within the Mathematical and Physical Sciences Directorate contribute funds to this award. This project performs research and education in high performance computing in the domain of nanotechnology. The five principle investigators at two universities, plus graduate student researchers and postdoctoral researchers will be engaged in atomistic simulations of entire nanoscale device elements and structures. The project will produce modeling tools that will be shared freely as a set of open source quantum simulation tools for petascale supercomputers in the broad area of nanoscience and nanotechnology. The proposed work consists of interdependent parts that together to produce robust, high performance petascale tools for quantum simulations at nanoscale. The development of such tools requires interdisciplinary, synergistic research in (i) methodology and implementation of quantum methods, (ii) profiling, performance modeling and automatic optimization of kernels, and (iii) algorithm development and tuning. The tools will be based on existing real-space multi-grid (RMG) method, which are now well-established and successfully applied to a large number of systems. Profiling and performance modeling are required to ensure effective utilization of the petascale hardware and will be accomplished by using tools and improving tools from three projects: PAPI, KOJAK and TAU. The remaining activity, algorithm development, focuses on implicit iterative methods and preconditioners that would improve convergence while preserving linear or nearly linear scaling, on variable step size implicit methods for quantum molecular dynamics, and on sparse eigensolvers with reduced scaling. Development of true petascale simulation tools at the quantum level is a substantial achievement with lasting intellectual impact. New research avenues will likely emerge, as results of such simulations are analyzed and new generally applicable concepts are created. Access to the open-source quantum simulation tools for petascale computers will have a broad impact. Advances in this field are relevant to virtually every area of scientific endeavor including chemistry, biochemistry and molecular biosciences, computer science and engineering, earth sciences, engineering, environmental sciences and materials science. Students and post doctoral researchers trained in this area will have significant opportunities for advancement and contributing impact on their own.NON-TECHNICAL SUMMARY:This award is made on a proposal submitted to the PetaApps Solicitation. The Office of Cyberinfrastructure, the Computer and Information Science and Engineering Directorate, and the Divisions of Materials Research, Physics, and Chemistry and the Office of Multidisciplinary Activities within the Mathematical and Physical Sciences Directorate contribute funds to this award. This project carries out research and education in high performance computing in the domain of nanotechnology. The work produces simulations of ultra-miniaturized electronic and structural elements, nearly the size of a millionth of an inch, which are nano-sized components of nano-devices. Computer software is being developed that can build, in a virtual sense, such device on an atom by atom basis and simulate operation and predict characteristics of such nano-device components. Such basic modeling gives highly reliable design characteristics of Nano Materials and the Devices and Processes involved in making the devices.The major work of the project is carried out by an interdisciplinary team of theoretical physicists, computer scientists and device modeling specialists. The five principle investigators at two universities, plus graduate student research and postdoctoral researchers will be engaged in atomistic simulations of entire nanoscale device elements and structures. The project will produce modeling tools that will be shared freely as a set of open source petascale quantum simulation tools in the broad area of nanoscience and nanotechnology.

技术摘要：该奖项是根据提交给PetaApps征求的提案颁发的。网络基础设施办公室、计算机和信息科学与工程理事会、材料研究、物理和化学部门以及数学和物理科学理事会内的多学科活动办公室为该奖项提供资金。该项目在纳米技术领域的高性能计算方面进行研究和教育。两所大学的五名主要研究人员，加上研究生研究人员和博士后研究人员，将从事整个纳米级器件元件和结构的原子模拟。该项目将产生建模工具，这些工具将作为一组开源量子模拟工具，免费共享给纳米科学和纳米技术广泛领域的千万亿次超级计算机。所提出的工作由相互依赖的部分组成，这些部分共同为纳米尺度的量子模拟提供强大、高性能的千兆级工具。这些工具的开发需要在以下方面进行跨学科的协同研究：(i)量子方法的方法论和实施，（ii）分析，性能建模和核的自动优化，以及（iii）算法开发和调优。该工具将基于现有的实空间多网格（RMG）方法，该方法现已建立并成功应用于大量系统。为了确保千兆级硬件的有效利用，需要分析和性能建模，并将通过使用工具和改进来自三个项目的工具来完成：PAPI， KOJAK和TAU。剩下的活动，算法开发，集中在隐式迭代方法和预条件，将提高收敛性，同时保持线性或近线性缩放，在可变步长隐式方法的量子分子动力学，和稀疏特征求解与减少缩放。在量子水平上开发真正的千万亿次模拟工具是一项具有持久智力影响的重大成就。随着模拟结果的分析和新的普遍适用的概念的产生，新的研究途径可能会出现。访问开源量子模拟工具的千万亿级计算机将产生广泛的影响。这一领域的进展与化学、生物化学和分子生物科学、计算机科学和工程、地球科学、工程、环境科学和材料科学等几乎所有科学领域都有关系。在这一领域接受培训的学生和博士后研究人员将有很大的晋升机会，并对他们自己产生影响。非技术摘要：该奖项是根据提交给PetaApps征集的提案颁发的。网络基础设施办公室、计算机和信息科学与工程理事会、材料研究、物理和化学部门以及数学和物理科学理事会内的多学科活动办公室为该奖项提供资金。该项目在纳米技术领域进行高性能计算的研究和教育。这项工作产生了超小型化电子和结构元件的模拟，其尺寸接近百万分之一英寸，是纳米器件的纳米级组件。正在开发的计算机软件可以在虚拟的意义上，在一个原子一个原子的基础上构建这样的设备，并模拟操作和预测这种纳米设备组件的特性。这样的基本建模为纳米材料以及制造这些器件所涉及的器件和工艺提供了高度可靠的设计特性。该项目的主要工作是由理论物理学家、计算机科学家和设备建模专家组成的跨学科团队进行的。两所大学的五名主要研究人员，加上研究生和博士后研究人员，将从事整个纳米级器件元件和结构的原子模拟。该项目将产生建模工具，这些工具将作为一组开源的千万亿次量子模拟工具，在纳米科学和纳米技术的广泛领域自由共享。