SAMOP 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
QI: Fachverband Quanteninformation
QI 4: Quantum Computing and Simulation (joint session Q/QI)
QI 4.1: Vortrag
Montag, 6. März 2023, 11:00–11:15, E214
An energy estimation benchmark for quantum computers — •Andreas J C Woitzik1, Edoardo Carnio1,2, and Andreas Buchleitner1,2 — 1Physikalisches Institut, Albert-Ludwigs-Universität Freiburg, Hermann-Herder-Straße 3, 79104 Freiburg im Breisgau, Federal Republic of Germany — 2EUCOR Centre for Quantum Science and Quantum Computing, Albert-Ludwigs-Universität Freiburg, Hermann-Herder-Straße 3, 79104 Freiburg im Breisgau, Federal Republic of Germany
While quantum-mechanical measurements yield intrinsically stochastic outcomes, the fluctuations in the output of current noisy intermediate-scale quantum (NISQ) devices are caused, for the large part, by imperfections in the hardware components and operations. We propose a simple energy estimation benchmark and use it to gauge noise-induced fluctuations in the output of IBM Quantum System One in Ehningen. We find that the errors we measure in our benchmark correlate only weakly with the reported calibration data of the machine. Moreover, a time-resolved analysis of the benchmark measure reveals periodic oscillations and unpredictable outliers that cannot be mildened by measurement error mitigation. We conclude that we cannot rely on single realizations of circuit outcomes, but rather on appropriately sampled ensembles.