We propose a novel error mitigation method that exponentially suppresses errors with a fixed number of qubits. Our method is feasible for the quantum advantage regime.
Error mitigation will be necessary for practical quantum advantage. Most error mitigation methods do not provide exponential suppression of errors. An exception is virtual distillation, but its large qubit requirements will likely prevent it from being useful for quantum advantage. Here, we propose a qubit-efficient virtual distillation method that can be implemented right now on existing hardware. This allows exponential suppression of errors with a fixed number of qubits.
For details see
“Qubit-efficient exponential suppression of errors”, P. Czarnik, A. Arrasmith, L. Cincio, P.J. Coles. arXiv:2102.06056