Table of Contents
LAB High Performance Optimization
--- Optimization Methods for Problems arising in Quantum Systems
| Lecturer | Prof. Dr. Petra Mutzel |
|---|---|
| Module | MA-INF 1222 |
| BASIS | Link2Basis |
| Type of Lecture | LAB |
| CP | 9 |
Application
This Lab has an upper limit on the number of participants. If you want to participate, please write an informal application e-mail to Petra Mutzel until March 31, 2025.
Content
We will focus on optimization methods arising in the context of Quantum Systems.
These could be, for example, structure formation problems in chemistry and theoretical physics, which are studied in the CRC 1639 NuMeriQS. Here, our subproject A05 (jointly with Prof. Barbara Kircher and Prof. Estela Suarez) aims at improving Ab initio molecular dynamics (AIMD) HPC-simulations and the analysis of their results for liquid systems in theoretical chemistry. The approach involves running multiple AIMD simulations simultaneously and using novel graph algorithms to exchange and analyze trajectory data. This will enable the study of more complex molecular systems and emergent structural properties. The goal is to make more efficient use of computer resources, improve application performance, and investigate complex chemical problems. The use of Modular Supercomputing Architecture (MSA) will match diverse workflow requirements and provide valuable insights for next-generation MSA-systems.
Another line of optimization methods studied in this Lab concern decomposition methods for solving Ising Spinglass problems that build the basis for parallel algorithms. An interesting candidate is the approach by Selby, which has shown to outperform quantum annealing systems for solution quality and computation time, such as, e.g., offered by D-Wave.
Within the lab, we will experimentally evaluate our approaches on benchmark data. For our evaluation we will make use of High Performance Systems, like Marvin at Bonn University or the HPC systems at the Jülich supercomputing center.
Requirements
Every participant is required to give a short presentation in the beginning of the Lab. A phase of literature study, algorithm design, and implementation follows. At the end, a presentation should be given (20-30 minutes per participant). We also expect a written report of about 10-12 pages of text per participant.