Tight Classical Speed Limits
Jean‐Charles Delvenne, Gianmaria Falasco · arXiv (Cornell University) · 2021
We propose a Classical Speed Limit for discrete state Markov chains. It establishes a lower bound on the expected number of jumps (activity) required by any time-varying Markovian protocol driving a state probability distribution to another in terms of the non-adiabatic (NA) entropy production along the path. We show that this Classical Speed Limit is close to tightness. We also discuss the connected problem of finding the time-varying protocol with minimum entropy production within a given expected number of jumps (which is solved by conservative forces, as recently proved). We show that these problems differ: in some cases, the protocol minimizing NA entropy production involves non-conservative forces. We also prove novel bounds for constant protocols.