Bounds on fluctuations of continuous machines in stochastic thermodynamics*

Thermal machines, such as heat engines and refrigerators, may be described at the nanoscale, with stochastic processes underlying the exchange of heat with thermal baths, and the input or output of work. As a result, heat and work are represented by stochastic variables, exhibiting fluctuations, which often play a significant role in evaluating device performance at this scale. In this talk, I will focus on continuous stochastic thermal machines: those whose operation is characterized by steady-state energy currents through an open system away from equilibrium. I will introduce basic models of such machines, and discuss a set of novel bounds on the ratio of fluctuations in their output currents to those in their input currents. Namely, for a given machine, this ratio is bounded from below by the square of the machine's efficiency, and from above by the square of the relevant Carnot bound. This leads to a tighter-than-Carnot bound on the efficiency itself. These results have been proven universally for continuous thermal machines operating near equilibrium, in the regime of linear response. Various analytic and computational approaches have been taken towards extending them to the far-from-equilibrium regime.