Quantum feedback enables finite-time Landauer erasure and near-unit efficiency in collective engines

Liu's team starts from the stochastic master equation for a monitored open quantum system and inserts a jump-feedback protocol that conditions the control on measurement outcomes. They obtain an exact finite-time relation linking the probability current, entropy production, and information gain, from which both a quantum speed limit and a power-efficiency trade-off follow directly. The derivation is analytic and holds for any Markovian jump process with instantaneous feedback.

In the information-erasure example the bound is saturated when feedback strength scales with system size; the collective engine reaches 98 percent information efficiency at nonzero power once N exceeds several hundred particles. These results rest on the assumption of perfect, instantaneous monitoring and feedback, conditions that remain experimentally demanding.

Related work on quantum Maxwell demons (Sagawa & Ueda, 2012) and finite-time thermodynamics (Deffner & Campbell, 2019) lacked the continuous-monitoring current term that Liu retains. The missing piece is a quantitative accounting of how feedback alters the dynamical contribution to dissipation, which the new relation supplies.

Next steps require embedding the protocol in a superconducting-qubit or trapped-ion platform to test whether measurement back-action and finite feedback latency preserve the predicted scaling. Without such tests the thermodynamic advantage remains formal.