Zero and negative energy dissipation at information-theoretic erasure
Laszlo B. Kish, Claes G. Granqvist, Sunil P. Khatri, Ferdinand Peper

TL;DR
This paper explores a novel information-theoretic approach to memory erasure that can be energy-neutral or even result in negative energy dissipation, challenging traditional thermodynamic limits like Landauer's principle.
Contribution
It introduces a new erasure scheme based on Shannon's binary channel formula that can achieve zero or negative energy dissipation, with potential implications for energy-efficient memory technologies.
Findings
Information-theoretic erasure can be energy-free in certain memory models.
A charge-based memory scheme can have negative energy dissipation during erasure.
Classical erasure processes can produce arbitrarily large negative energy dissipation.
Abstract
We introduce information-theoretic erasure based on Shannon's binary channel formula. It is pointed out that this type of erasure is a natural energy-dissipation-free way in which information is lost in double-potential-well memories, and it may be the reason why the brain can forget things effortlessly. We also demonstrate a new non-volatile, charge-based memory scheme wherein the erasure can be associated with even negative energy dissipation; this implies that the memory's environment is cooled during information erasure and contradicts Landauer's principle of erasure dissipation. On the other hand, writing new information into the memory always requires positive energy dissipation in our schemes. Finally, we show a simple system where even a classical erasure process yields negative energy dissipation of arbitrarily large energy.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsQuantum Computing Algorithms and Architecture · Advanced Memory and Neural Computing · Neural dynamics and brain function
