报告题目：Large-Scale Modeling of Complex Quantum Systems

报告人：袁声军 教授

报告时间：2022年4月1日（星期五）15：30-17：00

报告地点：在线报告（腾讯会议ID：354 146 031）

报告摘要：

A common approach to solving a quantum mechanics problem usually starts by looking for solutions to the time-independent Schrödinger equation. Observable quantities or physical properties are derived using eigenfunctions obtained from the diagonalization of Hamiltonian. However, as the computational costs  (memory and CPU time) scale cubically with the system size, this approach is unfavourable or inapplicable for many complex quantum problems. This talk will show new numerical methods for modelling quantum systems without any diagonalization. I will focus on condensed matter physics and introduce the tight-binding propagation method (TBPM). Unlike the standard approach, it is based on the numerical solution of the time-dependent Schrödinger equation, and the computational costs scale linearly with the system size. It has significant advantages in modelling large and complex quantum structures, ranging from mesoscopic to macroscopic level, without the requirement of any symmetry. I will briefly introduce the method and our simulation package TBPLaS (www.tbplas.net). At last, I will discuss our recent work on the self-consistent density functional theory (arXiv:2203.03465), which allow the large-scale modelling of millions of atoms from first principles.

报告人简介：

袁声军，武汉大学物理科学与技术学院教授，荷兰奈梅亨大学和北京计算科学研究中心客座教授。主要的研究领域为计算物理学、凝聚态理论和高性能科学计算，包括发展针对复杂量子体系的大尺度模拟方法，发展针对量子多体问题的计算方法，低维量子体系的电学、光学、输运和等离基元性质，通用量子计算机模拟，强关联自旋系统的弛豫和退相干等等。