Very interesting! I wonder that if they are expecting to use dissipative mechanisms in small system how would they be able to control thermal noise and reproducibility of the calculations?