SAMOP 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
QI: Fachverband Quanteninformation
QI 35: Quantum Computers: Algorithms and Benchmarking
QI 35.1: Vortrag
Freitag, 10. März 2023, 11:00–11:15, B305
Advantages of Measurement-based Variational Quantum Eigensolvers — •Anna Schroeder1,2, Matthias Heller3, and Mariami Gachechiladze2 — 1Merck KGaA, Frankfurter Str. 250, Darmstadt, Germany — 2Quantum Computing Group, TU Darmstadt, Mornewegstr. 30, Darmstadt, Germany — 3Fraunhofer IGD, Darmstadt, Germany
The variational quantum eigensolver (VQEs) is a hybrid algorithm to compute the lowest eigenvalue and its corresponding eigenvector for a given operator. The idea is to optimize classically over a parametrized quantum circuit, the ansatz, to generate a quantum state that minimizes the cost function, typically the expectation value of the Hamiltonian. Recently, a different approach to VQE has been considered. Ferguson et al., PRL 2021 discussed unifying the VQE framework with measurement-based quantum computing (MB-VQE). Here instead of parametrizing gates, one starts with highly entangled resource states (e.g., graph states) and optimizes over local measurements. This scheme has already demonstrated an advantage over the gate-based model for small perturbations of Toric code Hamiltonians, as it allows for more compact construction of certain ansaetze while enjoying shallower circuit depths - an imperative property for implementation on NISQ hardware. In our work, we deepen the investigation of MB-VQE advantage by considering more general resource states and larger classes of Hamiltonians, which helps us develop a more rigorous understanding of the advantageous ansaetze in MB-VQE for given problem classes.