SAMOP 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
QI: Fachverband Quanteninformation
QI 23: Poster II (joint session QI/Q)
QI 23.12: Poster
Mittwoch, 8. März 2023, 16:30–19:00, Empore Lichthof
Single qubit gate optimization based on ORBIT cost functions — •Catharina Broocks1,2, Max Werninghaus1, Niklas Glaser1,2, Federico Roy1,3, and Stefan Filipp1,2,4 — 1Walther-Meißner-Institut, Bayerische Akademie der Wissenschaften, Garching, Germany — 2Physik Department, Technische Universität München, Garching, Germany — 3Theoretical Physics, Saarland University, Saarbrücken, Germany — 4Munich Center for Quantum Science and Technologies (MCQST), Munich, Germany
Analytic control solutions for qubit gates are limited by the knowledge about modeled experimental hardware properties. To achieve high-fidelity gates for superconducting qubit devices, we optimize pulse parameters of analytic solutions with respect to experimental feedback loops. As cost function the ground state population after a net-identity series of Clifford gates is used. For small parameter sets, the parameter-landscape can serve as a reference to verify numerical system models and provide insight into the sensitivity and correlation of individual parameters. To find an optimal parameter configuration, we apply simultaneous multi-parameter optimization of single-qubit gates in form of CMA-ES closed-loop optimization. We analyze the behavior of the optimization algorithm when using features such as sensitivity adjustment, influence of various noise contributions and the design of the cost function to achieve a reliable and complete convergence of the algorithm. The optimization routine can then be used to verify and address various optimal control problems, such as robustness and avoidance of leakage out of the qubit subspace.