MC 5501
Speaker
Matthias Möller, Associate Professor in the Department of Applied Mathematics section Numerical Analysis at TU Delft
Title
Opportunities of Quantum Computing for Computational Science and Engineering
Abstract
Quantum computing (QC) is an emerging compute technology that has the potential to radically change the way we will be solving computational problems in the future. The potential power of QC stems from the exploitation of quantum mechanical principles, namely, superposition of states, quantum entanglement and quantum parallelism, which makes it possible to solve certain types of computational problems up to exponentially faster or with exponentially less memory storage.
In this talk we will give an overview of the related research activities at TU Delft and discuss several promising applications of quantum or quantum-assisted solution strategies in the Computational Science and Engineering domain. The first one is a quantum lattice Boltzmann method (QLBM) that aims at enabling large-scale fluid flow simulations on fault-tolerant gate-based quantum computers [1,2,3]. We will discuss the pros and cons of different encodings of the primary variables into the quantum register and provide executable quantum circuits for the three essential components - streaming, collision, and boundary treatment. In particular, we introduce a novel space-time encoding [2] that for the first time enables the realization of streaming and collision as unitary operators and prove that with conventional encodings (e.g., in the amplitudes or using basis state encoding) only one of the two steps can be realized as a unitary operator at a time. We finally present numerical results obtained on quantum computer emulators with our open-source QLBM software framework [4].
In the second part of the talk we present a novel approach to solve the nonlinear system of equations steming from an AC power flow model with the aid of quantum and digital annealing [5]. Next to discussing the derivation of the QUBO/Ising model formulation we show a first set of numerical results obtained with the help of D-Waves' Advantage System and Fujitsu's Digital Annealer. Our approach is particularly useful for ill-conditioned problem instances, where classical solution techniques, e.g., based on the Newton-Raphson method fail to converge. If time permits we will briefly consider another application of quantum annealing to a common problem in seismic imaging - stack-power maximization for refraction residual statics estimation [6].
References
[1] Merel A. Schalkers and Matthias Möller. Efficient and fail-safe quantum algorithm for the transport equation. Journal of Computational Physics, 502:112816, April 2024.
[2] Merel A. Schalkers and Matthias Möller. On the importance of data encoding in quantum Boltzmann methods. Quantum Information Processing, 23(1), January 2024.
[3] Merel A. Schalkers and Matthias Möller. Momentum exchange method for quantum Boltzmann methods, arXiv:2404.17618, April 2024.
[4] Calin Georgescu, Merel A. Schalkers, and Matthias Möller. QLBM - A quantum lattice Boltzmann software framework. In preparation.
[5] Zeynab Kaseb, Matthias Möller, Pedro P. Vergara, and Peter Palensky. Adiabatic Quantum Power Flow, Preprint [https://doi.org/10.21203/rs.3.rs-4368636/v1], May 2024.
[6] Marcin Dukalski, Diego Rovetta, Stan van der Linde, Matthias Möller, Niels Neumann, Frank Phillipson. Quantum computer-assisted global optimization in geophysics illustrated with stack-power maximization for refraction residual statics estimation. Geophysics 88 (2), February 2023.