Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

MemComputing vs. Quantum Computing

memcpu.com

Post reply on HN
Print view View original

1–10 of 18 posts

Re: MemComputing vs. Quantum Computing

#2

Post by generalizations »

I didn't know that memcomputing had already been solved. That's fantastic news...here's hoping it doesn't remain ridiculously expensive.

I take issue with this line, however.

> Both MemComputing and quantum computers are physics-based approaches to computation. That is, they set aside many of the foundations of traditional computer science and rethink computation from the ground up. For example, they don’t follow the von Neumann architecture, the computing framework that separates the processing unit from the memory unit, which is employed by all of today’s computers.

From what I recall of Von Neumann, his titular architecture was the 'first draft' of what he was going for, and we've been stuck with it before he could figure out version two. He based his design on the notion of cellular automaton, where of course the processing units are not separated from the memory units.

Rather, from what I recall, the Architecture we've been stuck with was just his best attempt at emulating what's here called memcomputing, but with the limited engineering resources at his disposal. If he hadn't died, I think he would have been trying to do proper memcomputing in the second draft. Only took us the better part of a century to catch up!

Re: MemComputing vs. Quantum Computing

#4

Post by bdamm »

generalizations wrote: post #2

I didn't know that memcomputing had already been solved. That's fantastic news...here's hoping it doesn't remain ridiculously expensive. I take issue with this line, however. > Both MemComputing and quantum computers are physics-based approaches to computation. That is, they set aside many of the foundations of traditional computer science and rethink computation from the ground up. For example, they don’t follow the…

I view the von Neumann architecture as more of a reference. When you get down to it, there's so much hardware trickery at play that it isn't easy to cleanly say that any specific computer fits precisely into what Von Neumann was describing. Consider multi-layer caches, CPUs that execute directly out of RAM, mapped IO, or even micro machine code design where the instruction set of the CPU is really an emulated layer on top of a deeper processor, and it's really hard to reconcile these modern machines as a Von Neumann architecture. I'd say that computer design has taken all the pragmatic approaches available. We have had wholesale architecture reboots every ten years or so. Even modern x86 doesn't resemble the original all that much.

However, as a conceptual bundle to teach aspiring computer science students about hardware? It's great!

Requiring that our software be able to simulate a Turing machine has probably held us back more, although that basic abstraction did get us to this point so in no way am I besmirching the work of Dr. Turing.

Re: MemComputing vs. Quantum Computing

#5

Post by inasio »

Their tagline at the end:

```MemComputing is available today and is delivering the performance expected of quantum computers across a wide range of valuable optimization problems in industry. ```

Quantum computers unfortunately are not expected to do well against traditional OR algorithms in most (all?) OR problems (Grover search only helps in brute force search; Shor it not quite an OR application, and relies on quantum effects), so not really a strong selling point in my view

Re: MemComputing vs. Quantum Computing

#7

Post by hn_throawlles »

inasio wrote: post #3

These guys claimed that they could factor semi-primes in polynomial time [0], essentially P == NP. Scott Aaronson predictably had some thoughts on that [1] [0] https://pubs.aip.org/aip/cha/article/27/2/023107/135054/Poly... [1] https://scottaaronson.blog/?p=2212

> “smuggle the exponentiality” somewhere that isn’t being explicitly considered, exactly like all proposals for perpetual-motion machines smuggle the entropy increase somewhere that isn’t being explicitly considered.

this is the gist of this. so he is saying that the scalability issues they will run into, come from the entropy increasing with the N such that heat is the real problem? (ah, so maybe this is also why Quantum computers are usually frozen???)

also, how big of an N are we talking about?

I see a weird echo here, I am thinking about exporting the waste (heat, byproducts, the specific 'how-to? by which, in a way, the entropy rules get 'enforced') as being the problem.

so then, he wrote:

> while the proponents wield the sharp steel of accepted physical law

I find this analogy quite cutting... with sharp enough steel and enough acceptance (regardless of how the acceptance comes about.... ok), the heat/entropy can get forced around? time to go get my tin foil shield,

then again, I have only been trying to understand exponentiation ever since I found out it existed, and I had not even heard about the "extended" version of the church-turing thesis, hence I'm some kind of ignorant crank (sorry if this comes across as bitter, but I am studying a master's degree in theory of computer science and I had not ever heard about an extended version of that famous thesis;l surely my own fault for not going to study to the USA or europe)

Re: MemComputing vs. Quantum Computing

#9

Post by marcodiego »

> Somewhat surprisingly, just the emulation of this circuit in software (Virtual Memcomputing Machine) provides solutions to these problems orders of magnitude faster and more accurately than today’s state of the art technologies.

Looks like market speech to me.

Re: MemComputing vs. Quantum Computing

#10

Post by ljlolel »

marcodiego wrote: post #9

> Somewhat surprisingly, just the emulation of this circuit in software (Virtual Memcomputing Machine) provides solutions to these problems orders of magnitude faster and more accurately than today’s state of the art technologies. Looks like market speech to me.

Yea that can’t be true. Otherwise you would be able to systematically modify an existing algorithm to a new one to mirror the Memcomputing method and create new state of the art algorithms that are faster by orders of magnitude? And somehow that hasn’t already been discovered and published and used preferentially to the published state of the art? Nonsense.
Post reply on HN