# Thermodynamic Computing: Better than Quantum? | Guillaume Verdon and Trevor McCourt, Extropic

**URL:** <https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317>\
**Category:** Machine Learning\
**Created:** [April 14, 2024, 11:32am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317 "2024-04-14T11:32:44Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![Peter2](https://avatars.discourse-cdn.com/v4/letter/p/d2c977/32.png) [@Peter2](https://discourse.numenta.org/u/Peter2)\
**Post date:** [April 14, 2024, 11:32am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/1 "2024-04-14T11:32:44Z")

</div>

[![](https://img.youtube.com/vi/OwDWOtFNsKQ/hqdefault.jpg "Thermodynamic Computing: Better than Quantum? | Guillaume Verdon and Trevor McCourt, Extropic") ](https://www.youtube.com/watch?v=OwDWOtFNsKQ)

---

<div class="post-metadata">

**Author:** ![cezar\_t](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/cezar_t/32/4451_2.png) [@cezar\_t](https://discourse.numenta.org/u/cezar_t)\
**Post date:** [April 17, 2024, 7:28am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/2 "2024-04-17T07:28:57Z")

</div>

Any kind of summary or reference to know whether it is worth 1 hour of time?  
Because when you consider the current stage of “quantum” any computing is better than it.

---

<div class="post-metadata">

**Author:** ![Peter2](https://avatars.discourse-cdn.com/v4/letter/p/d2c977/32.png) [@Peter2](https://discourse.numenta.org/u/Peter2)\
**Post date:** [April 18, 2024, 3:33am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/3 "2024-04-18T03:33:14Z")

</div>

They agree with you and they have a prototype chip.  
And without being sure about what the prototype hardware and software they have developed, and without being an insider or expert I agree with them and you about the whimpering future of the quantum computer hypers.

---

<div class="post-metadata">

**Author:** ![cezar\_t](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/cezar_t/32/4451_2.png) [@cezar\_t](https://discourse.numenta.org/u/cezar_t)\
**Post date:** [April 20, 2024, 12:03pm UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/4 "2024-04-20T12:03:59Z")

</div>

Since we all agree quantum is trash, let’s stick to the matter:

What “thermodynamic computing” means? What does it does and how? I won’t watch an hour video if it doesn’t start with answering these basic questions.  
An engineering/science podcast shouldn’t be a game-of-thrones mystery series “wait till you see who wins the battle” with “click like and subscribe before you-re too bored to resist any longer!”

---

<div class="post-metadata">

**Author:** ![dmac](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/dmac/32/5842_2.png) [@dmac](https://discourse.numenta.org/u/dmac)\
**Post date:** [April 21, 2024, 12:57am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/5 "2024-04-21T00:57:23Z")

</div>

As far as I can tell: They’re making an AI accelerator chip that uses analog physics (as opposed to digital transistors).

Their device will only be capable of running [energy-based models](https://en.wikipedia.org/wiki/Energy-based_model).

> **[Ushering in the Thermodynamic Future - Litepaper](https://www.extropic.ai/future)**
>
> We are very excited to finally share a bit more about what we are building: a full-stack hardware platform to harness the natural fluctuations of matter as a computational resource for Generative AI.

---

<div class="post-metadata">

**Author:** ![dmac](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.numenta.org/dmac/32/5842_2.png) [@dmac](https://discourse.numenta.org/u/dmac)\
**Post date:** [April 21, 2024, 1:22am UTC](https://discourse.numenta.org/t/thermodynamic-computing-better-than-quantum-guillaume-verdon-and-trevor-mccourt-extropic/11317/6 "2024-04-21T01:22:52Z")

</div>

On quantum computers:

Timestamp: 1:03:50

> Whats interesting about the physically inspired computing space is that every different kind of deivce has its own kind of natural set of algorithms that it can accererate, right? Because building a physics-based accelerator means your embedding some kind of math that you want the answer to in the dynamics of an analog system, right? And so when you build dramatically different devices, maybe like a quantum computer is really good at solving Schrodingers equation, or like a memristor array type of thing is really good at simulating memrisistor arrays. Our computer is really good at sampling from programmable distributions. So in a sense the point I’m trying to get at here is that there is room for a lot of different plays in thes space because every accelerator ends up being good at something different.