Linear-Scaling Electronic Structure and Quantum Simulations via the Stochastic Propagation Method

by Prof. Shengjun Yuan(袁声军) (Wuhan University)

Thursday 5 Nov 2026, 10:30 → 11:30 Asia/Shanghai

Tsung-Dao Lee Institute/N6F-N601 - Meeting Room (Tsung-Dao Lee Institute)

Host: Prof. Xin Liu  

Venue: TDLI Meeting Room N601

Tencent Meeting link: https://meeting.tencent.com/dm/9Aecpehs5DOo

Meeting ID: 570864770, no password

 

Abstract:

Electronic structure and quantum simulations in condensed matter physics often rely on solving the time-independent Schrödinger equation, with states and observables obtained via matrix diagonalization. However, the cubic scaling with system size severely limits the applicability of these approaches to complex, disordered, or large-scale systems. In this talk, I will introduce the stochastic propagation method (sPM), a linear-scaling approach that maps static problems onto the time-dependent Schrödinger equation framework. By propagating random states that sample the Hilbert space, this approach circumvents costly matrix diagonalization and enables calculations of electronic structure, optical response, quantum transport, magnetism, and plasmons at large length scales, expanding accessible simulation sizes by several orders of magnitude. Originally developed for large-scale tight-binding models, this methodology also extends to density functional theory. Furthermore, I will highlight extensions of sPM to quantum spin systems, the Hubbard model, and lattice dynamics (phonons), and conclude with recent advances in integrating stochastic propagation with machine learning.

 

Biography:

Shengjun Yuan is a Professor in the School of Physics and Technology at Wuhan University. He received his BSc from Zhejiang University in 2001 and his PhD from the University of Groningen, the Netherlands, in 2008. He has received several recognitions, including the Netherlands Foundation for Fundamental Research on Matter, China’s National Young Talent Program, and the National Science Fund for Distinguished Young Scholars. His research focuses on the development of novel computational methods in physics. He has developed large-scale first-principles approaches, including TBPM and DFPM, and applied them to a wide range of problems in complex quantum systems and condensed matter physics. He has also independently developed several computational physics software packages, such as TBPLaS and ABPLaS. Professor Yuan has published over 160 papers in leading international journals, including Nature, Science, and Nature Materials, with more than 80 articles in the Physical Review series.