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

The Case Against Quantum Computing

spectrum.ieee.org

Post reply on HN
Print view View original

31–40 of 89 posts

Re: The Case Against Quantum Computing

#31

Post by klodolph »

TheOtherHobbes wrote: post #5
zeroxfe wrote: post #2

Had to read to the bottom to get to the (really weak) "case against Quantum Computing": > I believe that, appearances to the contrary, the quantum computing fervor is nearing its end. That’s because a few decades is the maximum lifetime of any big bubble in technology or science. After a certain period, too many unfulfilled promises have been made, and anyone who has been following the topic starts to get annoyed by…

The technical argument is absolutely clear: quantum computing cannot work because it relies on manipulating and measuring an absolutely astronomical number of continuous variables with near-infinite precision. The argument may or may not be correct, but it deserves something more thoughtful than a dismissive response that doesn't even recognise the basic point the author is making.

The argument that quantum computing relies on manipulating continuous variables with "near-infinite precision" is flatly incorrect, if my understanding of the threshold theorem is correct. So no, quantum computers do not require extreme amounts of precision or extremely low error rates because it is possible to correct errors by making the computer larger.

https://en.wikipedia.org/wiki/Quantum_threshold_theorem

Re: The Case Against Quantum Computing

#32

Post by mathgenius »

This is a lame article, bordering on click-bait.

It is totally reasonable to predict a quantum computing winter, but all the supporting arguments here seem to be about how "hard-headed engineers" know best.

> So the number of continuous parameters describing the state of such a useful quantum computer at any given moment must be at least 2^1000, which is to say about 10^300. That’s a very big number indeed.

Yes this is the whole point: if it wasn't this big, we could just simulate it on a regular computer.

> To repeat: A useful quantum computer needs to process a set of continuous parameters that is larger than the number of subatomic particles in the observable universe.

Yep.

> it’s absolutely unimaginable how to keep errors under control for the 10^300 continuous parameters that must be processed by a useful quantum computer.

I do imagine it, every day. But I'm a professional, so maybe that doesn't count?

> How many physical qubits would be required for each logical qubit? No one really knows...

There are plenty of such estimates around. Obviously it depends on the noise rate of each qubit.

> all of the assumptions that theorists make about the preparation of qubits into a given state, the operation of the quantum gates, the reliability of the measurements, and so forth, cannot be fulfilled exactly. They can only be approached with some limited precision. So, the real question is: What precision is required? ... Amazingly, not only are there no clear answers to these crucial questions, but they were never even discussed!

There are boatloads of papers (and conferences) that discuss exactly this. Obviously it is a vital question to answer!

Re: The Case Against Quantum Computing

#33

Post by WillPostForFood »

tarboreus wrote: post #11
simonh wrote: post #8

Earlier quoted context omitted.

If something sounds too good to be true, it probably is.

Sulfa. White swans do come along, they're just incredibly rare.

Are they rare, or do they just happen over longer timeframes than we want? Just about any mainstream technology today would appear to be too good to be true to someone ~100 years ago. Air travel, television, computers, phones.

Re: The Case Against Quantum Computing

#34

Post by mathgenius »

TheOtherHobbes wrote: post #5
zeroxfe wrote: post #2

Had to read to the bottom to get to the (really weak) "case against Quantum Computing": > I believe that, appearances to the contrary, the quantum computing fervor is nearing its end. That’s because a few decades is the maximum lifetime of any big bubble in technology or science. After a certain period, too many unfulfilled promises have been made, and anyone who has been following the topic starts to get annoyed by…

The technical argument is absolutely clear: quantum computing cannot work because it relies on manipulating and measuring an absolutely astronomical number of continuous variables with near-infinite precision. The argument may or may not be correct, but it deserves something more thoughtful than a dismissive response that doesn't even recognise the basic point the author is making.

> it relies on manipulating and measuring an absolutely astronomical number of

"Manipulating" yes, that is the whole point. But you don't measure all those astronomical number of variables. The measurements stay sane, the state does not.

> it deserves something more thoughtful than a dismissive response

To me this article is itself a dismissive response to quantum computing.

Re: The Case Against Quantum Computing

#35

Post by tezthenerd »

Here is a way to see the fallacy of the OMG, its 10^300 variables, thats crazy style of “argument”.

Consider a probabilistic classical algorithm on 500 bits. Perhaps a Monte Carlo simulation of an Ising model for example.

Note first that the most general probability distribution over the 500 classical bits takes 2^500 real numbers to specify. (You have to specify P(000…0) and P(000…1) and… P(111…1)). [You should compare this to the 2^501 real parameters it takes to specify the quantum state of 500 qubits.]

To generate the most general such distribution perhaps you are restricted to using circuits where the gates act on at most n-bits at a time. Each gate can be described by a 2^n x 2^n bistochastic matrix comprised of (n-1)^2 real parameters. [You should compare this to the n^2 real parameters it takes to specify a 2^n x 2^n unitary matrix for a quantum gate acting on n qubits.]

Obviously its nuts to imagine you can generate all classical probability distributions over the 2^500 real parameters, particularly if you’re so mad as to think you are going to do it using a circuit comprised only of these n-bit gates!

Therefore useful classical monte carlo computing is obviously decades away.

(Oh and please trust me, I'm well know in stuff that isn't quantum computing.)

Re: The Case Against Quantum Computing

#36

Post by pavelrub »

sp332 wrote: post #30
pavelrub wrote: post #20

Earlier quoted context omitted.

The "quantum computers" we have are classical computers with very weird hardware. The thing everybody is trying to achieve is quantum supremacy - this hasn't been achieved or demonstrated yet, and might never will be.

They are not classical computers. We've had actual quantum logic gates since 1995. https://en.wikipedia.org/wiki/Timeline_of_quantum_computing And as linked above, IBM is renting out time on real quantum computers up to 17 qubits now. Intel, Google, and IBM had all announced plans for 49 or 50 qubit computers, which would have achieved quantum supremacy at the time, but IBM improved their simulator to 56 qubits last…

They are classical computers in the sense they are are no stronger than classical computers - they provide a different physical implementation, but computationally can't achieve anything that a classical computer can't. Everything those "quantum computers" do can be simulated with a classical computer. The entire motivation behind the development of QC is supremacy, and the existence of those "quantum computers" alone doesn't provide any proof that supremacy is possible - so it is definitely not just a question of scale.

Google already has a 72 qubit processor (Bristlecone), but whether they or anybody will be able to demonstrate supremacy in the current, next, or some future generation is entirely conjectural. The main limit is error correction, and there are theoretical arguments (Gil Kalai) for why achieving supremacy might be impossible.

Re: The Case Against Quantum Computing

#37

Post by amelius »

I'm wondering what will be first: practical fusion reactors or practical QC.

Also, will we at some point be able to linearly convert energy into compute power. In that case: would we still need QC?

(You might say that this conversion is already possible, but it requires human interaction)

Re: The Case Against Quantum Computing

#38

Post by hazeii »

Animats wrote: post #28
setquk wrote: post #4

Earlier quoted context omitted.

Sounds like Fusion. That’s still going on after 60 odd years.

We know that fusion is possible. We have H-bombs. We have various things that do small scale fusion. Getting out more power than is put in remains way out of reach.

I take the point, while noting H-bombs put out more power than is put in. Perhaps "getting more useful power out..."?

Re: The Case Against Quantum Computing

#39

Post by atq2119 »

tezthenerd wrote: post #35

Here is a way to see the fallacy of the OMG, its 10^300 variables, thats crazy style of “argument”. Consider a probabilistic classical algorithm on 500 bits. Perhaps a Monte Carlo simulation of an Ising model for example. Note first that the most general probability distribution over the 500 classical bits takes 2^500 real numbers to specify. (You have to specify P(000…0) and P(000…1) and… P(111…1)). [You should comp…

Monte Carlo simulation doesn't generate all classical probability distributions though. Attempting to do that would be nuts. Monte Carlo simulation only samples from a single, fixed distribution (or maybe a small number of distribution).

I'm not super convinced by the argument based on number of parameters either, but your analogy doesn't refute it at all.

Re: The Case Against Quantum Computing

#40

Post by core-questions »

krastanov wrote: post #15
TheOtherHobbes wrote: post #5

Earlier quoted context omitted.

The technical argument is absolutely clear: quantum computing cannot work because it relies on manipulating and measuring an absolutely astronomical number of continuous variables with near-infinite precision. The argument may or may not be correct, but it deserves something more thoughtful than a dismissive response that doesn't even recognise the basic point the author is making.

This is one of the standard complaints (the gist of it being that quantum computing is some form of analog computing, i.e. requiring near-infinite precision). For researchers in the field it becomes rather frustrating to have to repeat the same response without being heard, so I can understand the annoyance expressed in the parent comment. For what is worth, here is a good explanation of how this complaint misreprese…

> the gist of it being that quantum computing is some form of analog computing

It absolutely is - at least, in the only practical, real-today, working instantiation of it which is in the form of quantum annealing.

Post reply on HN