The goal here is to keep a bit of information $b$ at the temperature $T$ during a time $\delta t$.
I don't know if this question makes a lot of sense, considering the diversity of ways to encode information.
But is there a way to generalize all this techniques in term of state and particle?
Then could we obtain a lower bound on the energy we need to spend to keep this information unaltered.
It's a question very open, so don't hesitate to reformulate the problem in a way understandable for you. And I hope it's not a off-topic.