[oliveremberton]

Interesting refutation by IBM here:

In the preprint, it is argued that their device reached “quantum supremacy” and that “a state-of-the-art supercomputer would require approximately 10,000 years to perform the equivalent task.” We argue that an ideal simulation of the same task can be performed on a classical system in 2.5 days and with far greater fidelity.

https://www.ibm.com/blogs/research/2019/10/on-quantum-suprem...

[phh]

IMO the whole IBM blog is interesting, your quote isn't just a summary.

Basically what IBM is claiming is that Google's circuit doesn't do anything useful really, it is just meant to be very complicated to do in traditional computing systems. And even in this idealized position, Google's quantum computer doesn't transform an unsolvable problem into a solvable one.

According to IBM, "Quantum Supremacy" means accomplishing something useful to the society, that simply wouldn't be possible to do otherwise. Google's article doesn't show any sign of that, IBM claim.

[afthonos]

"Quantum Supremacy" is a technical term, not a colloquial one. It refers to showing that there exists a problem that a real, physical quantum computer can solve quickly that a classical computer cannot.

Reading the IBM article, they are fully aware of what "quantum supremacy" means in a technical sense, and they are urging the media not to use that term, since it will be misunderstood by the public. Their claim that Google has failed to achieve supremacy rests solely on their claim that they can simulate the circuit far faster (and scale the simulation linearly) using better classical algorithms.

That's a strong claim, and I'm interested in seeing what Google responds with.

Disclosure: I work at Google, but hahaha, no, I'm not cool enough to work on this.

[tagrun]

Physicists here. Note that they say it scales linearly in circuit depth (which is a trivial fact, and has always been true for classical simulations of quantum computers which are optimal in that regard --in fact, that is the case when doing it in the most naive way), not the number of qubits which is the quantum speed up referred in "quantum supremacy".

Another thing, this is actually Martinis' decades long work. I know Google recently started raining money down on his lab a couple years back, helping with the classical aspects, design etc, and media loves reporting as Google's Quantum Computer, but the actual quantum computer, the nitty gritty physics isn't Google's work. Martinis already had a working setup with ~10 qubits when Google started supporting him ~5 years ago.

This IBM "rebuttal" sounds a bit like cheating on multiple aspects, and the timing of the announcement is interesting. Note that they don't tell you how the memory requirements grow with the number of qubits either (which is exponential as well). I expect the response will be new toy computation proposals which will also be prohibitively expensive in classical memory (not just classical CPU with limited memory) in current supercomputers as well. If the experimentalists can roll out more qubits faster though (less likely), the "concern" will be addressed as well.

[afthonos]

Thanks for the detailed response; I am not a physicist, so I didn’t catch the sleight of hand in the linear scaling claim. The timing of the “rebuttal” is almost certainly intentional, and possible because of the accidental pre-publication last month. I hope the rebuttal is indeed specious, because it’s an exciting advance; I’m sure time will tell.

[tagrun]

It's pretty easy and requires no physics, actually.

Here's a simple, extended version.

A quantum gate is, mathematically speaking, a matrix. For a given physical system, of fixed number of qubits, obtaining that matrix on a classical computer takes (on average) a fixed amount of time, let's say T seconds. A quantum "circuit" is a sequence of quantum gates, applied consecutively in time, and you simply multiply them all to get their overall effect.

So if your circuit is made of 10 gates, the total CPU time is 10T, plus the time for 10-1 matrix multiplications. If it is 20 gates, then it is 20T plus time for 20-1 matrix multiplications.

Since multiplying two matrices of the same dimension also takes a fixed amount, on average, the simulation time grows linearly with circuit depth.

The quantum supremacy is related to how T grows as you increase the number of qubits, n (which is exponential, it's a 2^n by 2^n matrix).

[sanxiyn]

No, if you read Google paper, quantum supremacy is entirely about circuit depth scaling.

[tagrun]

No idea where exactly you got that idea (feel free to quote any part of the paper), but no, it isn't.

Even the brute force "simulation" of a quantum computer is like UN...U2.U1 where Us are unitarity matrices. The hard part is obtaining those unitaries (whose dimensions grow exponentially with the number of qubits). For fixed number of qubits though, once you have N unitaries, you do N matrix multiplications. If you double N, it'll take twice long on a classical and roughly twice on a quantum computer (different gates take different amount of time to implement). But on an actual quantum computer, there are tricks you can do (if the Hamiltonian allows) which may allow you to do it in fewer unitaries.

Circuit depth is still important because it is important 1) for modelling the noise in the device and extracting gate fidelities, that's basically how randomized benchmarking works although they're doing something else for fidelity estimates it still is a function of circuit depth 2) for doing anything meaningful when using a given set of basic building block gates.

[littlestymaar]

From the blog post (I haven't read the paper yet) that's also what I understand

> we ran random hard circuits with 53 qubits and increasing depth, until reaching the point where classical simulation became infeasible

Or am I misunderstanding something? (ELI5 plz, I'm not well-versed in quantum computing).

[sanxiyn]

Exact quote from the paper: "algorithm becomes exponentially more computationally expensive with increasing circuit depth". See also figure 4b, where circuit depth scaling is graphed.

[sanxiyn]

Thank you. I am not sure why people are talking about 2.5 days, since the key claim is "linearly scaling". If simulation is linearly scaling in this regime, the entire proof is indeed questionable.

[mazesc]

IBMs response seems strong regarding "linear scaling", but could they give a complexity-theoretic argument rather than an empirical one?

[sanxiyn]

I think IBM didn't elaborate because it is kind of obvious if you are into this, and if you are not, linearly scaling graph from 10 to 30 is more than good enough. But since we have niche audience here, let's elaborate.

Quantum supremacy is O(n) claim. Then what is n? The answer is that there are two parameters, not one. In Google paper, n is number of qubits and m is circuit depth. n is well known to be difficult to increase. m is not easy either, because if your qubit isn't stable enough, you can't run deep circuit. What Google did is to run n=53 and m=20.

Then, why do you need n=53 and m=20? After all, you could see whether it's exponential by, say, trying n and m from 10 to 15, it doesn't need to take days and years. The answer is that there is time-space tradeoff available and if your n is constant, exp(n) space (but still constant space, since n is constant) poly(m) time algorithm is available, and if your m is constant, exp(m) space (but still constant space, since m is constant) poly(n) time algorithm is available. So if you want to show exponential speedup, you need to be able to exclude these algorithms by increasing n or m enough such that exp(n) or exp(m) space is not realizable. Google chose n=53 such that exp(n) is not realizable, and ran scaling experiment on m.

This is what they mean whey they say in the paper "Up to 43 qubits, we use a Schrödinger algorithm, which simulates the evolution of the full quantum state... Above this size, there is not enough random access memory (RAM) to store the quantum state". Now what IBM is saying is becoming clear. There is not enough RAM, but there is enough disk, so exp(n) where n=53 is realizable, and simulation runs linear in m. It's not a new algorithm, it's exactly the same algorithm Google ran up to n=43. So 43 qubits clearly can't demonstrate quantum supremacy. For the same reason, 53 qubits can't either.

[mannykannot]

Thanks for this clear explanation of the issue! This looks to me like a convincing rebuttal of the specific claim Google made about classical runtime at this particular size, but does it rule out Google claiming supremacy by using the problem constrained such that n == m? Or would Google's device have exponential runtime growth in that variant?

[htfu]

Thanks for the breakdown! But so in other words, it’s still not all that far off that (realistic amounts of) disk storage can’t contain the state and scaling reverts to how they assert it.

[mazesc]

Thanks a lot, that was insightful!

[skywhopper]

But from the Google blog it sounds like the “problem” they chose boils down to “emulating a quantum computer”. Which doesn’t sound like it’s exactly in the spirit of the test, since the quantum computer can obviously emulate itself—and with zero wasted operations!

[afthonos]

It's exactly the spirit of the test; the missing element is that the circuit they're emulating can scale in qubits. So the question is, if you add one cubit, what happens? Can classical computers keep up? The tests were run on a range of cubits with 50+ being the largest number, which is where the 10,000 year claim comes from. Anything further than that just isn't feasible to compute on classical computers, because they don't scale like that.

So why is this in the spirit of the test? Because this means that there are some problems that can only be efficiently solved by quantum computers. So this establishes "supremacy" in the sense that while a quantum computer can efficiently solve any classical computing problem, a classical computer cannot solve any quantum computing problem.

The distance between having this proof-of-concept and having meaningful speedups on real problems that cannot be matched by classical computers is very large; all this tells you is that it's possible, and looking more might not be a waste.

[marktangotango]

> Disclosure: I work at Google, but hahaha, no, I'm not cool enough to work on this.

I often wonder why people feel compelled to write this. There nothing in your profile or in your post history to prove unequivocally you do or don’t work anywhere in particular so why even mention it?

Edit; responders are citing ethics and company policy; but does anyone really think vague hand wavey speculation on a public message board is relavant to anthing?

Sure if you work in the ads group and you posted "wow, our division is in trouble, competitor Z is really killing it, and our quarter earnings are going to miss big time!" Yeah that matters, but GP? Really?

[otterley]

First, it’s a professional ethics requirement. If there is the possibility of a conflict of interest in our statement, we need to disclose it to avoid misleading anyone.

In addition, there are plenty of ways to figure out where an HN member is employed besides looking at their post history. It’s not necessarily found in publicly available data, but it can be found.

Also, you have to account for future acts. Disclosing early can help insure against accusations about misconduct if you were to out yourself — or someone were to out you — later.

[afthonos]

Company policy. This prevents headlines like "Google had employees secretly astroturf to give traction to their quantum supremacy claims" if someone decides to truly dig. I theoretically could pretend I don't work here, but I find life is easier if I just follow reasonable rules.

(I did forget to mention I didn't speak for Google, but it should be pretty obvious from context)

[AYBABTME]

To disclose possible bias, in this case maybe parent feels like their opinion might be biased toward Google's claim.

[smadurange]

I agree. Possibly to just give a bit more weight to what they say. So far, I've mostly seen this with Google employees here. Can't remember seeing anyone else mentioning this except when promoting a product.

1. Regarding policy, I think a simple disclosure like this is my personal opinion not that of my employer would be good enough. 2. Indicate possible bias: nope, one should always understand any opinion is biased due to many different reasons. So, unless this is a rigorous analysis of something, this is quite uncalled for.

[IanCal]

> According to IBM, "Quantum Supremacy" means accomplishing something useful to the society, that simply wouldn't be possible to do otherwise. Google's article doesn't show any sign of that, IBM claim.

I thought that the whole point was to show that what you'd done definitely wasn't just build a complex classical processor. The usefulness of the calculation is besides the point.

[lisptw102019]

Both of you are slightly off.

Quantum Supremacy is more about where quantum computers are clearly superior to non-quantum computers to solve a problem.

[clSTophEjUdRanu]

It's not about doing useful work it's about finding an actual workload that cannot be solved faster classically.

[hos234]

Will just point out it's 25 years since Deep Blue and 10 since Watson/Jeopardy. And boy were lot of claims made back then. Good ol'IBM has yet to deliver anything.

[onetimemanytime]

Regardless of what they say, Quantum computer solved it in 200 seconds and IBM states that their super-computer solves it in 2.5 days (Google says 10,000 years.) Even if we take IBM's claim at face value, 200 seconds vs 2.5 DAYS is still a lot of improvement, semantics aside. Not to mention that quantum computers have just started

[sanxiyn]

No, 2.5 days is irrelevant. This is about proving exponential speedup, but IBM is claiming linear scaling, so it is a serious challenge.

[mmusson]

They say linear in time for increasing circuit depth not increasing qubits. They also don’t address the classical algorithm being exponential in memory as well.

[mda]

What if it is linear but angle is so steep that you require all the sands in the world to achieve parity with a classical computer, does it still count?

[sanxiyn]

Of course. Quantum supremacy is a claim about computational complexity. 200 seconds vs 2.5 days is 1000x slowdown. In practical terms, x^10 is even worse than 1000x (linear), but if quantum computer could be simulated in x^10, that would definitely disprove quantum supremacy.

[drcode]

I think in that case you would probably be able to formulate the problem in a different way to show that the angle arises through an exponential process (or more) and that the problem isn't truly linear.

[contravariant]

Maybe but 1) that's hard to prove and 2) it's hard to achieve unless you're looking at some sub-problem where the quantum and classical algorithm actually do scale differently.

[ridewinter]

One way of viewing Google's achievement is as a result that cannot be explained without invoking parallel universes. That's pretty amazing.

I posted an article explaining more: https://news.ycombinator.com/item?id=21337739

[mtgx]

All of that is true, but according to IBM's own definition of "quantum advantage" I think Google at least achieved that.

All in all, it seems more of a battle of semantics, and the two companies talking past each other. It's kind of how everyone seems to interpret Moore's Law to mean "any progress in computational performance whatsoever" these days, as opposed to the original definition of "transistors doubling in the same space every 18-24 months," and then using that to "prove" that "Moore's Law is still alive." (it's not, and hasn't been for a long time).

I think in the end what's important here is that it was proven that a quantum computer can do "something" (yes, even anything at all is a big milestone) significantly faster than a classical system.

Because this is what matters in the end that quantum computers can do some tasks significantly faster than classical systems. They don't have to do only tasks that are "impossible" for classical systems to be useful. It's also why we use GPUs, FPGAs, and ASICs for some tasks, instead of just using CPUs for everything.

I do agree that it was in Google's interest not to work so hard on optimizing the classical system for the simulation, to make its quantum computer look that much more impressive. But if the quantum computer is still faster, then it doesn't really matter.

[sanxiyn]

No, if your standard is "faster", even D-Wave achieved that. Quantum supremacy is about exponential speedup, and in light of IBM's claim, Google's claim of quantum supremacy is very much in doubt.

[mellosouls]

I haven't processed the blog, but considering IBMs resources wouldn't it be simpler to refute it by actually doing it in 2.5 days?

Or is there some catch that undermines that headline claim?

[svantana]

That's 2.5 days on the 13 MW Summit supercomputer, for a single run. Presumably you would need multiple runs to get good statistics.

[photon-torpedo]

No, the classical simulation should give you the whole wavefunction, from which you can get the full probability distribution. One run should be enough.

[ehsankia]

Does it scale linearly? Is it 5 days on a half as good super computer? They just need to show that it can be done in any reasonable finite amount of time, not 2.5 days.

[earenndil]

The main bottleneck is memory/storage capacity, not speed. Beyond that (if you have the requisite storage capacity and bandwidth), yes: if you have half the FLOPS and double the time, you get the same result.

[sipjca]

Is this the case? I thought that parallel speedup is not linear. so it might actually not be 2x slower but maybe 1.5x or something depending on IPC overhead and all that

[onion2k]

I think the quote from IBM means it'd take 2.5 days to run. That doesn't include actually implementing it.

[forgot-my-pw]

That might be a pretty expensive run on a supercomputer

[hamilyon2]

The margin between 200 seconds and 2.5 days still exists

[photon-torpedo]

As far as I understand, the 2.5 days classical simulation gives you the total wavefunction, from which you can read the probability distribution. You only need to run it once. It's not clear to me whether the 200 seconds quantum computation is for getting (or rather sampling) the probability distribution, or for just one measurement. My point is, the classical simulation actually gives you much more data than the quantum computation -- you get the actual full probabilty distribution, not just an approximated sample of it (guess that's why IBM claims "higher fidelity").

[reportgunner]

I thought quantum supremacy was about having a quantum computer capable of doing more operations per second than any binary computer can do

Do you think it wouldn't be possible to simply upgrade the supercomputer to shave some time off those 2.5 days ?

To me it seems clear that everyone gullible already bought into the Cloud, ML/AI, CI/CD buzzwords and now they need a new buzzword to market.

[comicjk]

Quantum supremacy is about scaling (big-O notation) - taking a classically O(2^N) problem and making it polynomial. The empirical experiments are just to show that the scaling works, and the uncertainty comes from the fact that we still can't fit very large N problems into current quantum computers. Effectively, they don't have enough RAM to even load the problem (i.e. enough available quantum states).

[enkid]

That's not quantum supremacy. It's about doing something on a quantum computer that would be infeasible on a standard computer. This is because in some scenarios a single calculation by a quantum computer could be equivalent to a vast number (many trillions) of calculations made by a standard computer.

[reportgunner]

> more operations per second than any binary computer can do

> doing something on a quantum computer that would be infeasible on a standard computer

Sorry, but I fail to see how these two are different.

[lippel82]

They use different "algorithms", so solving a problem can be done in very different ways for a quantum and a classical computer. Comparing the number of operations per second does not make sense because the advantage of quantum algorithms lies in the fact, that they can achieve the same results with (exponentially) fewer operations.

[reportgunner]

We can keep moving the goalposts and I can rephrase my claim to:

> it's a state when a quantum computer can achieve results that would require more operations on a binary computer than a binary computer has proven to be able to do

but the point will still stand.

We can't say we achieved quantum supremacy for this one thing because binary computers still have supremacy over everyting else.

I guess we can agree here that quantum supremacy was definitely not achieved since we are not clear on the definition of said quantum supremacy.

[hoseja]

I think it's about running algorithms with big-O complexities lower than can be ran on a classical computer, not actually more ops.

[ewalk153]

I think that’s where the

> and with far greater fidelity

comes into play.

[the8472]

A physical system will always be the fastest implementation of a computer simulating its own physics.

Some overhead is expected when using different hardware to simulate it.

What matters is asymptotic computational and memory complexity.

[lallysingh]

Yes but nobody cares.

The QC isn't cheap or available enough to be relevant is your can compute it reasonably well in a few days.

[clSTophEjUdRanu]

I think IBM is full of shit and badly needs this PR boost that is about to be snatched away by Google.

[excalibur]

I don't think I need to invoke superposition to conceive of them both being full of shit at the same time.

[make3]

do you have any substance to back this claim up

[dang]

There's also a separate thread discussing IBM's post: https://news.ycombinator.com/item?id=21333105.