Linear Scaling Electronic Structure Methods

线性缩放电子结构方法

• 批准号: 9618323
• 负责人: Gustavo Scuseria
• 金额: $ 24.8万
• 依托单位: William Marsh Rice University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-03-15 至 2000-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9618323&HistoricalAwards=false
• 关键词: Linear; Scaling; Electronic; Structure; Methods

The Theoretical and Computational Chemistry program is supporting professor Scuseria at Rice University. During the next three years he will continue his work on Linear scaling in Electronic structure methods. The research contains both computational development, as well as applications. The computational work is focused in at five distinct areas. These include (1) Improvements to the Fast Multipole method, (2) Improvements to the exchange-correlation quadratures, (3) diagonalization bottlenecks (4) O(N) Hartree-Fock algorithms, and finally (5) improving DFT capabilities. Specific chemical problems to be addressed in the next funding period will emphasize Fullerene growth, graphite sheets and carbon nanotubes. First principles electronic structure calculations of molecules remain limited to systems of modest size because of the steep computational cost. The objective of this research is to develop fast, high accuracy methods for the prediction of properties of large systems such as biomolecules and nonostructures. The quest here is to develop full linear scaling methods. Work will continue in this area with specific emphasis on removing the existing computational bottlenecks.

理论和计算化学计划支持 莱斯大学的Scuseria教授。 在接下来的三年里，他 将继续他的工作在电子结构方法的线性标度。 这项研究既包括计算开发， 应用. 计算工作集中在五个不同的 地区 其中包括（1）对快速多极子方法的改进，（2） 交换相关积分的改进;（3）对角化 瓶颈（4）O（N）Hartree-Fock算法，最后（5）改进 DFT能力。 下一步要解决的具体化学问题 融资期间将强调富勒烯的增长，石墨片和 碳纳米管 分子电子结构的第一性原理计算 由于陡峭的计算速度， 成本 本研究的目标是开发快速、高精度的 预测大型系统性能的方法， 生物分子和非结构。 这里的追求是发展全面的 线性尺度方法 将继续在这一领域开展工作， 重点是消除现有的计算瓶颈。