MRI: Acquisition of a Network Cluster of Advanced Workstations for First Principles Electronic Structure Calculations of Complex Materials

MRI：获取先进工作站网络集群，用于复杂材料的第一原理电子结构计算

• 批准号: 0116068
• 负责人: Gabriel Kotliar
• 金额: $ 22.41万
• 依托单位: Rutgers University New Brunswick
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-09-01 至 2003-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0116068&HistoricalAwards=false
• 关键词: MRI; Acquisition; Network; Cluster; Advanced

With this award from the Major Research Instrumentation program, Rugters University will acquire instrumentation for the setup of a network of clusters of advanced workstations. This system will be used for first-principles investigations of physical properties of complex materials of technological and scientific importance. The ultimate goal is to be able to predict the physical properties of solids with the same accuracy and reliability that one can achieve for small molecules using quantum-chemistry methods. These computations will implement novel, state-of-the-art algorithms for electronic-structure calculations in a massively parallel setting. The techniques span three general areas: (i) Kohn-Sham density-functional theory in a plane-wave pseudopotential formulation, for static calculations of systems for which electronic correlations are weak or of intermediate strength; (ii) Standard extensions of density functional theory and dynamical mean-field theory for strongly-correlated electron systems; and (iii) effective-Hamiltonian statistical methods for bridging from these microscopic methods to finite-temperature simulations.The materials systems to be investigated span a wide range, including high-K dielectrics (mostly Zr, Hf, Ta, Nb, Si, and Y oxides and oxynitrides), systems exhibiting complex martensitic transitions, ferroelectrics, oxides and selenides, Yb- and Ce-based heavy-fermion compounds, actinides, high temperature superconductors, and oxygen and hydrogen absorbed on W and other transition-metal surfaces. The research will involve the training of students and postdoctoral researches in computational materials science.With this award from the Major Research Instrumentation program, Rutgers University will acquire instrumentation to build an appropriate system on the basis of low cost high performance Linux PC's. These will be used in research on complex materials and carry calculations requiring high-speed parallel computer platforms. Complex materials, with tunable properties that can be controlled and optimized for a given application, will continue to play a role of ever-increasing importance in the new century. As our understanding of the laws that govern complex materials improves, and as computer technology and computational methods advance together, our ability to design and modify such materials will have an ever-expanding impact on science and technology. Computational condensed-matter science plays a very important role in this quest by providing new approaches to the modeling of material properties and by shedding light on the unusual properties of materials via simulations at the atomistic scale. In addition to this scientific objectives, the project will have a significant impact on the education and training of graduate and undergraduate students and postdoctoral researchers in computational materials science at Rutgers University.

凭借主要研究仪器计划的这一奖项，罗格斯大学将获得用于建立高级工作站集群网络的仪器。该系统将用于具有技术和科学重要性的复杂材料的物理特性的第一性原理研究。最终目标是能够以与使用量子化学方法预测小分子相同的准确性和可靠性来预测固体的物理性质。这些计算将在大规模并行设置中实现新颖、最先进的电子结构计算算法。这些技术涵盖三个一般领域：（i）平面波赝势公式中的 Kohn-Sham 密度泛函理论，用于电子相关性较弱或中等强度的系统的静态计算； (ii) 强相关电子系统的密度泛函理论和动态平均场理论的标准扩展； （iii）有效的哈密顿统计方法，用于从这些微观方法到有限温度模拟。要研究的材料系统范围很广，包括高 K 电介质（主要是 Zr、Hf、Ta、Nb、Si 和 Y 氧化物和氮氧化物）、表现出复杂马氏体转变的系统、铁电体、氧化物和硒化物、Yb-和 Ce基重费米子化合物、锕系元素、高温超导体以及吸附在W和其他过渡金属表面上的氧和氢。该研究将涉及计算材料科学方面的学生和博士后研究的培训。凭借重大研究仪器计划的这一奖项，罗格斯大学将获得仪器，以在低成本高性能 Linux PC 的基础上构建适当的系统。这些将用于复杂材料的研究并进行需要高速并行计算机平台的计算。具有可针对特定应用进行控制和优化的可调特性的复杂材料将在新世纪继续发挥越来越重要的作用。随着我们对复杂材料规律的理解不断加深，以及计算机技术和计算方法的共同进步，我们设计和修改此类材料的能力将对科学和技术产生不断扩大的影响。计算凝聚态科学在这一探索中发挥着非常重要的作用，它提供了材料特性建模的新方法，并通过原子尺度的模拟揭示了材料的不寻常特性。除了这一科学目标外，该项目还将对罗格斯大学计算材料科学领域的研究生和本科生以及博士后研究人员的教育和培训产生重大影响。