The topic of reversible computers and backwards execution is quite different to logical reversibility. My experiences of reversibility were all driven by correctness in software and it didn’t matter that under the bonnet nothing truly executes backwards. Now AI with its hungry datacentres has made energy a top priority problem to solve, and development of reversibile hardware to achieve thermodynamic reversibility seems feasible.

Logical and thermodynamic reversibility

Large Language Models are subject to the laws of physics in a bad way, because they use so much power and make so much heat. With reversibility, physics might come to our rescue and greatly reduce the amount of power required.

This all starts with Landauer’s 1961 principle that energy is not consumed by computation but by the erasure of information. Erasure forces a two-state system into one state which increases entropy, which exits as heat. Bennett’s 1973 extension showed that any computation can be restructured to erase nothing, so that the inputs are recoverable from outputs at every step, something he called logical reversibility, and which carries no Landauer penalty. From this we can see thermodynamic reversibility is the physical consequence of logical reversibility: when no information is erased, no entropy is generated, and the energy used to perform each computational step can be recovered and reused rather than lost as heat. And if you run an AI datacentre that is what you need.

The dependency runs in one direction only: thermodynamic reversibility requires logical reversibility, because an erased bit commits an irrecoverable entropy debt before the hardware gets any say.

My old friend reversible execution doesn’t even get to play here, because it works in almost the opposite direction thermodynamically. Record-and-replay reversibility functions by doing a lot of extra logging so that the system can be wound back. Logging involves writing information to memory, and in the case of overwriting that means destroying information, which costs energy according to Landauer. The original forward computation still destroys information in all the usual ways, and the reply is just more of that again. It is a different category altogether, but the name is confusing and the spirit of what is being done is similar.

Logical reversibility

Janus is a reversible language developed in 1982 and formally specified in 2007. Janus makes it impossible to write a program that discards information, and provides a program inverter that runs any Janus program cleanly backwards without a history tape. The harder problem is extending this to concurrent programs, where interleaving makes reversal non-trivial.

A 2022 mathematical paper on Reversing an Imperative Concurrent Programming Language from the University of Leicester demonstrates this diffficult problem is solvable. The paper Reversible Execution for Robustness in Embodied AI and Industrial Robots (one author in common with the previous paper) says:

We thus demonstrate how a traditional AI-based planning approach is enriched by an underlying reversible execution model that relies on the embodiment of the robot system

In 2025 it was shown that reversible architectures can work at scale for Large Language Models where transformer architectures become reversible. This works by reconstructing hidden states during backpropagation rather than storing them, and achieves order-of-magnitude memory reductions without sacrificing accuracy. This is not yet running on reversible hardware because such hardware doesn’t exist yet, but it demonstrates that the software side may be solvable. The models train on conventional GPUs, but they are logically reversible.

Taking this paper at face value, it would seem such logical reversibility meets the precondition for thermodynamic reversibility, giving us energy that can be claimed back if the hardware allows it. And that is where physics is on our side.

Thermodynamic reversibility

Vaire Computing in London are building practical reversible hardware which preseves the energy gains that logical reversibility permits to be reclaimed, described in their 2025 hardware paper. The underlying principle is to stop discarding information, so that you then stop paying the penalty for doing so. Their stated aim is to have a product in 2027.

It seems the Reversibile Computing 2026 conference in Torino has Vaire as its first commercial sponsor, suggesting the academic and engineering communities may be meeting in the middle.