The problem of quantifying the thermodynamic cost of bit erasure has received a great attention since its first proposal [1,2]. Its relation with previously considered problems of the thermodynamics of information as Maxwell’s demon and Szilárd’s engine has been extensively revisited and extended in the last decades. More recently, Landauer’s bound has been derived using the approach of stochastic thermodynamics (either classical or quantum) in terms of informational quantities [4,5]. In this framework, a classical or quantum bit is encoded in the state of a given system with very few degrees of freedom (most often one) whose external manipulation in the presence of a heat bath is designed to yield the logical state zero regardless of the initial state.
In their recently published work , Buffoni and Campisi demonstrate a simple and effective method to erase $N$ qubits at once. This is achieved exploiting the physics of spontaneous symmetry breaking of a many-body system composed of $N$ interacting qubits, namely, a quantum annealer. As the authors show, the interaction between the individual qubits and the symmetry breaking due to a phase transition lead to a cooperative many-body effect. Inspired by the proposal of Ref. , the authors implement a cyclic process that drives the entire system from a ferromagnetic phase to a paramagnetic one and back. In the first half of the cycle, the protocol traverses a critical point where the spontaneous symmetry breaking occurs.
The protocol proposed in Ref.  also suggests that one can use a macroscopic quantity such as the magnetization as a logic register. Buffoni and Campisi are the first to effectively confirm this prediction using remote access to a quantum annealer. Once gain, the key mechanism behind the erasure protocol is the spontaneous symmetry breaking that the macroscopic system undergoes. Its role can be more easily understood through the analogy suggested in Ref.  with the Szilárd engine. There, the insertion of the piston in the middle of the box can be interpreted as spontaneous symmetry breaking since it splits the phase space in two parts whose permutation leaves the Hamiltonian invariant.
Interestingly, Landauer’s principle remains unchanged when the register is a macroscopic quantity. It only requires two possible “values” (either positive or negative magnetization) to encode the two logical bits. Nevertheless, when this macroscopic register is reset, it can imply the resetting of its individual components as well. Buffoni and Campisi show that this is indeed the case when the quantum annealer is manipulated according to the protocol discussed above. Moreover, they show that this occurs with exceptional performance in terms of time duration, energy consumption (not including energy costs for operating the annealer or for the control operations to perform the protocol) and success rate.
Thus, the authors’ work opens the door for further exploration of this mechanism in other macroscopic systems with relevant impact on new mesoscale quantum technologies. Their results also leaves several interesting open questions regarding the performance achieved.
 M. Esposito, K. Lindenberg, and C. Van den Broeck, Entropy production as correlation between system and reservoir, New J. Phys. 12, 013013 (2010). https://doi.org/10.1088/1367-2630/12/1/013013.
 M. Esposito and C. Van den Broeck, Second law and Landauer principle far from equilibrium, Europhys. Lett. 95, 40004 (2011). https://doi.org/10.1209/0295-5075/95/40004.
 J. M. R. Parrondo, The Szilárd engine revisited: entropy, macroscopic randomness, and symmetry breaking phase transitions, Chaos 11, 725 (2001). https://doi.org/10.1063/1.1388006.
This View is published in Quantum Views under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. Copyright remains with the original copyright holders such as the authors or their institutions.