An Efficient User-Side Nulling Calibration for Quantum Annealing Computers

Abstract

(Withdrawn) This work describes an efficient user-side method of calibrating and correcting quantum annealing computers. For quantum annealing computers based on the Ising model, the method measures the residual bias of the h and J coefficients. Once measured, these biases can then be nulled in subsequent runs for any problem of interest. This method also returns a temperature for each qubit based on the measured versus the expected qubit distributions computed from a Boltzmann distribution model.

This preprint was withdrawn because the theoretical discussion of the inferred temperature of qubit couplers assumes an equilibrium distribution, which is not the case during the annealing cycles