Shtetl-Optimized

As 2015 winds down, I thought I’d continue my tradition of using this blog to showcase some awesome student projects from my graduate class.  (For the previous project showcases from Quantum Complexity Theory, see here, here, and here.  Also see here for the showcase from Philosophy and Theoretical Computer Science.)

This fall, I taught 6.S899, a one-time “Seminar on Physics and Computation” that focused on BosonSampling, complexity and quantum gravity, and universality of physical systems.  There were also lots of guest lectures and student presentations.  Unfortunately, we didn’t do any notes or recordings.

Fortunately, though, the students did do projects, which were literature reviews some of which ventured into original research, and all nine have agreed to share their project reports here!  So enjoy, thanks so much to the students for making it a great class, and happy holidays.

• Computational Complexity of Spectral Gaps, by Anand Natarajan.
• Further Extensions of Clifford Circuits and Their Classical Simulation Complexities, by Dax Koh.
• On the Complexity of Stoquastic Hamiltonians, by Ian Kivlichan.
• Gravitational Attacks on Relativistic Quantum Cryptography, by Jayson Lynch.
• Tensor Networks, Quantum Error Correction, and AdS/CFT, by John Napp.
• Computation in a Topological Quantum Field Theory, by Daniel Epelbaum and Raeez Lorgat.
• Quantum Complexity, Statistical Physics, and Combinatorial Optimization, by Rolando La Placa.
• Building and Bounding Quantum Bernoulli Factories, by Theodore Yoder.

Update (Dec. 23): Here are two conference announcements that I’ve been asked to make: Innovations in Theoretical Computer Science (ITCS) 2016, January 14-16 in Cambridge MA, and the Fifth Women in Theory Workshop, at the Simons Institute in Berkeley, May 22-25, 2016.

Follow

Just when I thought I could relax, as the waters slowly receded from the latest D-Tsunami, my inbox and Facebook feed once again lit up with inquiries—this time, asking me to confirm or deny that “A Paradox at the Heart of Mathematics Makes a Physics Problem Unanswerable.”

Uh-oh!

Luckily for my blood pressure, though, this one turned out to refer to something that more-or-less deserves the hype.  In particular, it’s about a phenomenal 146-page paper by Cubitt, Perez-Garcia, and Wolf, which just appeared this week in Nature (in condensed form, of course).  Incidentally, yeah, his name really is Toby Cubitt, pronounced like “qubit.”  He’s a good guy.

To those in quantum computing, Cubitt et al.’s breakthrough is old news, having already been on the arXiv for almost a year (we’ve also had a talk at MIT about it).  The arXiv has created a funny phenomenon, where you learn something new and cool, assimilate it, move on, and then a year later, everyone is suddenly asking you have you seen this thing, is it for real, etc. etc., just because the thing got some rubber stamp like acceptance to Nature that caused the press to pick it up.  Like, dude, I was into the undecidability of the spectral gap way before it went mainstream.

One more amusing anecdote before we dive into the math.  In his Nature News piece popularizing Cubitt et al.’s result, the writer Davide Castelvecchi quotes Rebecca Goldstein, the brilliant novelist and biographer of Kurt Gödel, as saying: “Turing thought more clearly about the relationship between physics and logic than Gödel did.”  Here’s what happened: Nature News wrote to Rebecca to ask what Gödel’s own thoughts were about the relation between undecidability and physics.  Rebecca passed the request along to me.  So I wrote back to her, arguing that they might just as well ask what Turing thought, since the Cubitt et al. result is “really” about Turing-undecidability (with Gödel-undecidability just an automatic corollary), and at any rate:

I also think that Turing thought more clearly about the relationship between logic and physics than Gödel did (indeed, Gödel himself said that it was only Turing‘s analysis of the notion of computability, in terms of actual physical machines that one could imagine building, that convinced him that computability had been properly defined).

Rebecca passed that back to Nature News, agreeing with it, and then at some point the quote became hers.  Far from being miffed about this, I consider having my forgettable words attributed to a genius like Rebecca to be one of the great honors of my life.  (By pure coincidence, she and I are having lunch next week; hopefully this will butter her up.)

So, OK, let me restate Cubitt et al.’s great theorem in less pop-sciencey terms than Nature News used.  (You could also just read the paper‘s intro, which is exceedingly clear, but what the hell—I’m here to serve.)

Suppose you have two-dimensional material made of a bunch of stationary particles, each with local Hilbert space dimension d, which are arranged on an L×L square grid (so, there are L² particles in all).  And suppose there’s some fixed d²-dimensional Hamiltonian h, with a local copy h_i,j=h acting on each neighboring pair of particles (i,j).  (I.e., the material is translationally invariant, with the same laws of physics acting throughout.)  Let H be the total Hamiltonian: that is, the sum of the h_i,j‘s over all the neighboring (i,j)’s.

Then a huge fraction of all of physics—quantum field theory, condensed-matter physics, you name it—can be summarized as, you’re trying to figure out the eigenvalues and eigenvectors of H.  The lowest eigenvalue, λ₀, tells you your material’s ground energy, while the higher eigenvalues, λ₁,λ₂,…, tell you the next discrete energy levels that the material can jump up to.  The corresponding eigenvectors tell you which quantum states the material is sitting in when it has these energies: the ground state v₀, and the excited states v₁,v₂,…  Those, in turn, determine basically everything you could want to know about the material: whether it superconducts, etc. etc.

Of course, the eigenvalues and eigenvectors will depend on the lattice size L.  Equally obviously, for any fixed L, you could in principle compute all the eigenvalues and eigenvectors by just diagonalizing some huge-ass matrix.  (That matrix being H.)  But physicists are usually more interested in the limiting behavior as L goes to infinity.  One of their most basic distinctions is: the material is gapped if λ₁-λ₀, the difference between the first excited energy and the ground energy, converges to some positive value or even grows with L as L→∞.  It’s gapless if λ₁-λ₀ converges to 0 as L→∞.  (Actually, Cubitt et al. use more technical definitions of both of these concepts, but we’ll ignore that.)

Cubitt et al.’s theorem now says the following: for some fixed, constant local dimension d, there is no algorithm that takes as input the local Hamiltonian h (say, as a d²×d² matrix of algebraic numbers), and that decides whether the material is gapped or gapless.  Indeed, you can reduce the halting problem to that problem, in such a way that the material will be gapped if your Turing machine halts, or gapless if it runs forever.

As an immediate corollary, there’s some 2D material—characterized by a translationally-invariant local Hamiltonian h on particles of local dimension d—such that whether the material is gapped or gapless is independent of the axioms of ZF set theory, or whatever else your favorite axioms might be.  (Proof: build a Turing machine M that halts if and only if it finds an inconsistency in set theory, then run Cubitt et al.’s reduction from the halting problem.  By Gödel, if set theory is consistent then it can’t prove whether M halts or not.)

Cubitt et al. never bother to work out the local dimension d that suffices for them, but it could be worked out, and it’s probably at least in the tens of thousands.  Thus, their result leaves open the possibility that there’s an algorithm to decide gaplessness for 2D lattices of qubits (i.e., the special case d=2), or other “reasonably low-dimensional” quantum systems.  We simply don’t know right now.  Another tantalizing open question is whether there’s an algorithm to decide gaplessness for one-dimensional spin chains—again, even in the special case d=2.  Right now, the best we have in that direction is a difficult recent result of Bravyi and Gosset, which gives an algorithm to decide gaplessness for one-dimensional, frustration-free chains of qubits.  (Here “frustration-free,” an amusing term that does not well describe this subject as a whole, means that you can minimize the energy H by minimizing the energies of each h_i,j individually.  Or, if you think of H as a SAT instance, it’s satisfiable.)

But while the exact value of d where uncomputability kicks in is still up for grabs, it’s extremely important that d is some fixed, universal constant, independent of the Turing machine.  Indeed, as Cubitt et al. point out in their paper, this is the only feature that makes their new result not a trivial corollary of the uncomputability of Wang tiling.  The latter is a famous result from 1966, which says that there’s no algorithm that takes as input a finite set of tiles, and that tells you whether, using unlimited copies of each tile, you could cover the entire plane (or equivalently, arbitrarily large finite regions of the plane).  I.e., this is yet another “natural” math problem that secretly encodes the halting problem.

The fact that d is fixed also means that, in order to encode larger and larger Turing machines into the local Hamiltonian h (as you must, if you want to embed the halting problem), you need to use more and more bits of precision (!) in the ~d⁴ real numbers that define h.  This then raises a question: how do you actually extract a description of a Turing machine from the binary expansions of the real numbers that define your Hamiltonian?  To do this, Cubitt et al. use Kitaev’s phase estimation algorithm—which, interestingly, is the only part of their construction that uses quantum mechanics in any way.  One thing that I’d love to understand better is whether the phase estimation is really essential here, or whether the analogous classical question, with the “Hamiltonian” given by a probability distribution over classical constraints, could also be proved to be undecidable for some fixed value of d—thereby showing that Cubitt et al.’s discovery had nothing to do with quantum mechanics.

(It’s possible that the answer to this is obvious; I didn’t think about it deeply.  Note that if the “classical Hamiltonian” is also deterministic, then the problem must be decidable for every fixed d, since there are only finitely many possible h’s, and we could cache all the answers in a lookup table.)

Anyway, it’s now my professional duty, as the prickly, curmudgeonly blogger I am, to end the post by shooing you away from two tempting misinterpretations of the Cubitt et al. result.

First, the result does not say—or even suggest—that there’s any real, finite physical system whose behavior is Gödel- or Turing-undecidable.  Thus, it gives no support to speculations like Roger Penrose’s, about “hypercomputing” that would exceed the capabilities of Turing machines.  The reason, again, is that as soon as you fix a lattice size L, everything becomes computable.  The Cubitt et al. result applies only to questions about the limiting behavior, as the number of particles goes to infinity.  But we already knew lots of examples of physical systems for which predicting their behavior in some infinite limit is at least as hard as the halting problem: for instance, the Wang tiles discussed earlier, or Post rewrite systems, or even Turing machines themselves.  Local Hamiltonians are a profound, nontrivial addition to that list—one that will be particularly striking to physicists, many of whom calculate the spectral gaps of at least 50 Hamiltonians between dinner and dessert.  But in some sense, there was no a-priori reason why a problem this general, about physical systems of unbounded size, ought to have been computable.

Second, the result does not say that any particular question physicists want an answer to—for example, the million-dollar Yang-Mills mass gap problem—is Gödel-undecidable.  “All it says,” is that the possibility that some real-world question of that kind could be undecidable isn’t totally closed off.  The Nature News piece stresses this latter implication a lot—as, admittedly, do Cubitt et al. themselves.  But to put things in perspective: four logicians proved around 1970 that there’s no algorithm to decide whether an arbitrary polynomial equation has an integer solution, thereby giving a negative solution to Hilbert’s Tenth Problem.  Yet with few exceptions, “working number theorists” barely even noticed this development, nor was (say) Andrew Wiles dissuaded from proving Fermat’s Last Theorem, by the absence of a general algorithm to do things like what he was trying to do.  (Indeed, the absence of a general algorithm was shown even earlier for equations like FLT, which have variables in the exponent.)  So I doubt the mathematical physicists who calculate spectral gaps for a living will be any more terrified than the number theorists were, to learn that they’ve been laboring their entire lives on the shores of the halting problem.  “Good for us, then!” they could rightly reply.  “Maybe our jobs won’t be so easy to automate.”

Update (Dec. 20): My colleague Seth Lloyd calls my attention to a PRL paper of his from 1993, which also discusses the construction of physical systems that are gapped if a given Turing machine halts and gapless if it runs forever.  So this basic idea has been around for a while.  As I explained in the post, the main contribution of the Cubitt et al. paper is just to get undecidability into “the sort of system physicists could plausibly care about” (or for which they could’ve plausibly hoped for an analytic solution): in this case, 2D translationally-invariant nearest-neighbor Hamiltonians with bounded local dimension.

Follow

Update (Dec. 16):  If you’re still following this, please check out an important comment by Alex Selby, the discoverer of Selby’s algorithm, which I discussed in the post.  Selby queries a few points in the Google paper: among other things, he disagrees with their explanation of why his classical algorithm works so well on D-Wave’s Chimera graph (and with their prediction that it should stop working for larger graphs), and he explains that Karmarkar-Karp is not the best known classical algorithm for the Number Partitioning problem.  He also questions whether simulated annealing is the benchmark against which everything should be compared (on the grounds that “everything else requires fine-tuning”), pointing out that SA itself typically requires lots of tuning to get it to work well.

Update (Dec. 11): MIT News now has a Q&A with me about the new Google paper. I’m really happy with how the Q&A turned out; people who had trouble understanding this blog post might find the Q&A easier. Thanks very much to Larry Hardesty for arranging it.

Meanwhile, I feel good that there seems to have been actual progress in the D-Wave debate! In previous rounds, I had disagreed vehemently with some of my MIT colleagues (like Ed Farhi and Peter Shor) about the best way to respond to D-Wave’s announcements. Today, though, at our weekly group meeting, there was almost no daylight between any of us. Partly, I’m sure, it’s that I’ve learned to express myself better; partly it’s that the “trigger” this time was a serious research paper by a group separate from D-Wave, rather than some trash-talking statement from Geordie Rose. But mostly it’s that, thanks to the Google group’s careful investigations, this time pretty much anyone who knows anything agrees about all the basic facts, as I laid them out in this blog post and in the Q&A. All that remains are some small differences in emotional attitude: e.g., how much of your time do you want to spend on a speculative, “dirty” approach to quantum computing (which is far ahead of everyone else in terms of engineering and systems integration, but which still shows no signs of an asymptotic speedup over the best classical algorithms, which is pretty unsurprising given theoretical expectations), at a time when the “clean” approaches might finally be closing in on the long-sought asymptotic quantum speedup?

Another Update: Daniel Lidar was nice enough to email me an important observation, and to give me permission to share it here.  Namely, the D-Wave 2X has a minimum annealing time of 20 microseconds.  Because of this, the observed running times for small instance sizes are artificially forced upward, making the growth rate in the machine’s running time look milder than it really is.  (Regular readers might remember that exactly the same issue plagued previous D-Wave vs. classical performance comparisons.)  Correcting this would certainly decrease the D-Wave 2X’s predicted speedup over simulated annealing, in extrapolations to larger numbers of qubits than have been tested so far (although Daniel doesn’t know by how much).  Daniel stresses that he’s not criticizing the Google paper, which explicitly mentions the minimum annealing time—just calling attention to something that deserves emphasis.

In retrospect, I should’ve been suspicious, when more than a year went by with no major D-Wave announcements that everyone wanted me to react to immediately. Could it really be that this debate was over—or not “over,” but where it always should’ve been, in the hands of experts who might disagree vehemently but are always careful to qualify speedup claims—thereby freeing up the erstwhile Chief D-Wave Skeptic for more “””rewarding””” projects, like charting a middle path through the Internet’s endless social justice wars?

Nope.

As many of you will have seen by now, on Monday a team at Google put out a major paper reporting new experiments on the D-Wave 2X machine.  (See also Hartmut Neven’s blog post about this.)  The predictable popularized version of the results—see for example here and here—is that the D-Wave 2X has now demonstrated a factor-of-100-million speedup over standard classical chips, thereby conclusively putting to rest the question of whether the device is “truly a quantum computer.”  In the comment sections of one my previous posts, D-Wave investor Steve Jurvetson even tried to erect a victory stele, by quoting Karl Popper about falsification.

In situations like this, the first thing I do is turn to Matthias Troyer, who’s arguably the planet’s most balanced, knowledgeable, trustworthy interpreter of quantum annealing experiments. Happily, in collaboration with Ilia Zintchenko and Ethan Brown, Matthias was generous enough to write a clear 3-page document putting the new results into context, and to give me permission to share it on this blog. From a purely scientific standpoint, my post could end right here, with a link to their document.

Then again, from a purely scientific standpoint, the post could’ve ended even earlier, with the link to the Google paper itself!  For this is not a case where the paper hides some crucial issue that the skeptics then need to ferret out.  On the contrary, the paper’s authors include some of the most careful people in the business, and the paper explains the caveats as clearly as one could ask.  In some sense, then, all that’s left for me or Matthias to do is to tell you what you’d learn if you read the paper!

So, OK, has the D-Wave 2X demonstrated a factor-10⁸ speedup or not?  Here’s the shortest answer that I think is non-misleading:

Yes, there’s a factor-10⁸ speedup that looks clearly asymptotic in nature, and there’s also a factor-10⁸ speedup over Quantum Monte Carlo. But the asymptotic speedup is only if you compare against simulated annealing, while the speedup over Quantum Monte Carlo is only constant-factor, not asymptotic. And in any case, both speedups disappear if you compare against other classical algorithms, like that of Alex Selby. Also, the constant-factor speedup probably has less to do with quantum mechanics than with the fact that D-Wave built extremely specialized hardware, which was then compared against a classical chip on the problem of simulating the specialized hardware itself (i.e., on Ising spin minimization instances with the topology of D-Wave’s Chimera graph). Thus, while there’s been genuine, interesting progress, it remains uncertain whether D-Wave’s approach will lead to speedups over the best known classical algorithms, let alone to speedups over the best known classical algorithms that are also asymptotic or also of practical importance. Indeed, all of these points also remain uncertain for quantum annealing as a whole.

To expand a bit, there are really three separate results in the Google paper:

1. The authors create Chimera instances with tall, thin energy barriers blocking the way to the global minimum, by exploiting the 8-qubit “clusters” that play such a central role in the Chimera graph.  In line with a 2002 theoretical prediction by Farhi, Goldstone, and Gutmann (a prediction we’ve often discussed on this blog), they then find that on these special instances, quantum annealing reaches the global minimum exponentially faster than classical simulated annealing, and that the D-Wave machine realizes this advantage.  As far as I’m concerned, this completely nails down the case for computationally-relevant collective quantum tunneling in the D-Wave machine, at least within the 8-qubit clusters.  On the other hand, the authors point out that there are other classical algorithms, like that of Selby (building on Hamze and de Freitas), which group together the 8-bit clusters into 256-valued mega-variables, and thereby get rid of the energy barrier that kills simulated annealing.  These classical algorithms are found empirically to outperform the D-Wave machine.  The authors also match the D-Wave machine’s asymptotic performance (though not the leading constant) using Quantum Monte Carlo, which (despite its name) is a classical algorithm often used to find quantum-mechanical ground states.
2. The authors make a case that the ability to tunnel past tall, thin energy barriers—i.e., the central advantage that quantum annealing has been shown to have over classical annealing—might be relevant to at least some real-world optimization problems.  They do this by studying a classic NP-hard problem called Number Partitioning, where you’re given a list of N positive integers, and your goal is to partition the integers into two subsets whose sums differ from each other by as little as possible.  Through numerical studies on classical computers, they find that quantum annealing (in the ideal case) and Quantum Monte Carlo should both outperform simulated annealing, by roughly equal amounts, on random instances of Number Partitioning.  Note that this part of the paper doesn’t involve any experiments on the D-Wave machine itself, so we don’t know whether calibration errors, encoding loss, etc. will kill the theoretical advantage over simulated annealing.  But even if not, this still wouldn’t yield a “true quantum speedup,” since (again) Quantum Monte Carlo is a perfectly-good classical algorithm, whose asymptotics match those of quantum annealing on these instances.
3. Finally, on the special Chimera instances with the tall, thin energy barriers, the authors find that the D-Wave 2X reaches the global optimum about 10⁸ times faster than Quantum Monte Carlo running on a single-core classical computer.  But, extremely interestingly, they also find that this speedup does not grow with problem size; instead it simply saturates at ~10⁸.  In other words, this is a constant-factor speedup rather than an asymptotic one.  Now, obviously, solving a problem “only” 100 million times faster (rather than asymptotically faster) can still have practical value!  But it’s crucial to remember that this constant-factor speedup is only observed for the Chimera instances—or in essence, for “the problem of simulating the D-Wave machine itself”!  If you wanted to solve something of practical importance, you’d first need to embed it into the Chimera graph, and it remains unclear whether any of the constant-factor speedup would survive that embedding.  In any case, while the paper isn’t explicit about this, I gather that the constant-factor speedup disappears when one compares against (e.g.) the Selby algorithm, rather than against QMC.

So then, what do I say to Steve Jurvetson?  I say—happily, not grudgingly!—that the new Google paper provides the clearest demonstration so far of a D-Wave device’s capabilities.  But then I remind him of all the worries the QC researchers had from the beginning about D-Wave’s whole approach: the absence of error-correction; the restriction to finite-temperature quantum annealing (moreover, using “stoquastic Hamiltonians”), for which we lack clear evidence for a quantum speedup; the rush for more qubits rather than better qubits.  And I say: not only do all these worries remain in force, they’ve been thrown into sharper relief than ever, now that many of the side issues have been dealt with.  The D-Wave 2X is a remarkable piece of engineering.  If it’s still not showing an asymptotic speedup over the best known classical algorithms—as the new Google paper clearly explains that it isn’t—then the reasons are not boring or trivial ones.  Rather, they seem related to fundamental design choices that D-Wave made over a decade ago.

The obvious question now is: can D-Wave improve its design, in order to get a speedup that’s asymptotic, and that holds against all classical algorithms (including QMC and Selby’s algorithm), and that survives the encoding of a “real-world” problem into the Chimera graph?  Well, maybe or maybe not.  The Google paper returns again and again to the subject of planned future improvements to the machine, and how they might clear the path to a “true” quantum speedup. Roughly speaking, if we rule out radical alterations to D-Wave’s approach, there are four main things one would want to try, to see if they helped:

1. Lower temperatures (and thus, longer qubit lifetimes, and smaller spectral gaps that can be safely gotten across without jumping up to an excited state).
2. Better calibration of the qubits and couplings (and thus, ability to encode a problem of interest, like the Number Partitioning problem mentioned earlier, to greater precision).
3. The ability to apply “non-stoquastic” Hamiltonians.  (D-Wave’s existing machines are all limited to stoquastic Hamiltonians, defined as Hamiltonians all of whose off-diagonal entries are real and non-positive.  While stoquastic Hamiltonians are easier from an engineering standpoint, they’re also the easiest kind to simulate classically, using algorithms like QMC—so much so that there’s no consensus on whether it’s even theoretically possible to get a true quantum speedup using stoquastic quantum annealing.  This is a subject of active research.)
4. Better connectivity among the qubits (thereby reducing the huge loss that comes from taking problems of practical interest, and encoding them in the Chimera graph).

(Note that “more qubits” is not on this list: if a “true quantum speedup” is possible at all with D-Wave’s approach, then the 1000+ qubits that they already have seem like more than enough to notice it.)

Anyway, these are all, of course, things D-Wave knows about and will be working on in the near future. As well they should! But to repeat: even if D-Wave makes all four of these improvements, we still have no idea whether they’ll see a true, asymptotic, Selby-resistant, encoding-resistant quantum speedup. We just can’t say for sure that they won’t see one.

In the meantime, while it’s sometimes easy to forget during blog-discussions, the field of experimental quantum computing is a proper superset of D-Wave, and things have gotten tremendously more exciting on many fronts within the last year or two.  In particular, the group of John Martinis at Google (Martinis is one of the coauthors of the Google paper) now has superconducting qubits with orders of magnitude better coherence times than D-Wave’s qubits, and has demonstrated rudimentary quantum error-correction on 9 of them.  They’re now talking about scaling up to ~40 super-high-quality qubits with controllable couplings—not in the remote future, but in, like, the next few years.  If and when they achieve that, I’m extremely optimistic that they’ll be able to show a clear quantum advantage for something (e.g., some BosonSampling-like sampling task), if not necessarily something of practical importance.  IBM Yorktown Heights, which I visited last week, is also working (with IARPA funding) on integrating superconducting qubits with many-microsecond coherence times.  Meanwhile, some of the top ion-trap groups, like Chris Monroe’s at the University of Maryland, are talking similarly big about what they expect to be able to do soon. The “academic approach” to QC—which one could summarize as “understand the qubits, control them, keep them alive, and only then try to scale them up”—is finally bearing some juicy fruit.

(At last week’s IBM conference, there was plenty of D-Wave discussion; how could there not be? But the physicists in attendance—I was almost the only computer scientist there—seemed much more interested in approaches that aim for longer-laster qubits, fault-tolerance, and a clear asymptotic speedup.)

I still have no idea when and if we’ll have a practical, universal, fault-tolerant QC, capable of factoring 10,000-digit numbers and so on.  But it’s now looking like only a matter of years until Gil Kalai, and the other quantum computing skeptics, will be forced to admit they were wrong—which was always the main application I cared about anyway!

So yeah, it’s a heady time for QC, with many things coming together faster than I’d expected (then again, it was always my personal rule to err on the side of caution, and thereby avoid contributing to runaway spirals of hype).  As we stagger ahead into this new world of computing—bravely, coherently, hopefully non-stoquastically, possibly fault-tolerantly—my goal on this blog will remain what it’s been for a decade: not to prognosticate, not to pick winners, but merely to try to understand and explain what has and hasn’t already been shown.

Update (Dec. 10): Some readers might be interested in an economic analysis of the D-Wave speedup by commenter Carl Shulman.

Another Update: Since apparently some people didn’t understand this post, here are some comments from a Y-Combinator thread about the post that might be helpful:

(1) [T]he conclusion of the Google paper is that we have probable evidence that with enough qubits and a big enough problem it will be faster for a very specific problem compared to a non-optimal classical algorithm (we have ones that are for sure better).

This probably sounds like a somewhat useless result (quantum computer beats B-team classical algorithm), but it is in fact interesting because D-Wave’s computers are designed to perform quantum annealing and they are comparing it to simulated annealing (the somewhat analogous classical algorithm). However they only found evidence of a constant (i.e. one that 4000 qubits wouldn’t help with) speed up (though a large one) compared to a somewhat better algorithm (Quantum Monte Carlo, which is ironically not a quantum algorithm), and they still can’t beat an even better classical algorithm (Selby’s) at all, even in a way that won’t scale.

Scott’s central thesis is that although it is possible there could be a turning point past 2000 qubits where the D-Wave will beat our best classical alternative, none of the data collected so far suggests that. So it’s possible that a 4000 qubit D-Wave machine will exhibit this trend, but there is no evidence of it (yet) from examining a 2000 qubit machine. Scott’s central gripe with D-Wave’s approach is that they don’t have any even pie-in-the-sky theoretical reason to expect this to happen, and scaling up quantum computers without breaking the entire process is much harder than for classical computers so making them even bigger doesn’t seem like a solution.

(2) DWave machines are NOT gate quantum computers; they call their machine quantum annealing machines. It is not known the complexity class of problems that can be solved efficiently by quantum annealing machines, or if that class is equivalent to classical machines.

The result shows that the DWave machine is asymptotically faster than the Simulated Annealing algorithm (yay!), which suggests that it is executing the Quantum Annealing algorithm. However, the paper also explicitly states that this does not mean that the Dwave machine is exhibiting a ‘quantum speedup’. To do this, they would need to show it to outperform the best known classical algorithm, which as the paper acknowledges, it does not.

What the paper does seem to be showing is that the machine in question is actually fundamentally quantum in nature; it’s just not clear yet that that the type of quantum computer it is is an improvement over classical ones.

(3) [I]t isn’t called out in the linked blog since by now Scott probably considers it basic background information, but D-Wave only solves a very particular problem, and it is both not entirely clear that it has a superior solution to that problem than a classical algorithm can obtain and it is not clear that encoding real problems into that problem will not end up costing you all of the gains itself. Really pragmatic applications are still a ways into the future. It’s hard to imagine what they might be when we’re still so early in the process, and still have no good idea what either the practical or theoretical limits are.

(4) The popular perception of quantum computers as “doing things in parallel” is very misleading. A quantum computer lets you perform computation on a superposed state while maintaining that superposition. But that only helps if the structure of the problem lets you somehow “cancel out” the incorrect results leaving you with the single correct one. [There’s hope for the world! –SA]

Follow

1. Emma Goldman, the fiery early-20^th-century anarchist, is credited for giving the world the immortal refrain “if I can’t dance, I don’t want to be part of your revolution” (actually it’s not clear that she ever said it so pithily, but she did express such a thought).  Admittedly, no one would mistake me for either a dancer or an anarchist, but I’ve always felt a kinship with Goldman over her terpsichorean line in the sand.  The other day, it occurred to me that there’s a parallel sentence that sums up my entire political philosophy—on the one hand, my default instinct to side with the downtrodden and with the progressive left, but on the other, my dissent from any even vaguely anti-STEM, anti-rationality, or anti-nerd undercurrents, and my refusal to join any popular uprising that seems liable (for example) to delay the discovery of a P≠NP proof, by inconveniencing the people working on one.

So, here’s my sentence, which you should feel free to reprint on t-shirts and coffee mugs as desired:

If I can’t do math, I don’t want to be part of your revolution.

2. Over at Scientific American‘s website, John Horgan posted an account of a workshop on Integrated Information Theory, which I attended a couple weeks ago at NYU (along with David Chalmers, Giulio Tononi, Christof Koch, Max Tegmark, and a dozen or so others).  I was the “official skeptic” of the workshop, and gave a talk based on my blog post The Unconscious Expander.  I don’t really agree with what Horgan says about physics and information in general, but I do (of course) join him in his skepticism of IIT, and he gives a pretty accurate summary of what people said at the workshop.  (Alas, my joke about my lunch not being poisoned completely bombed with the IIT crowd … as I should’ve predicted!)  The workshop itself was lots of fun; thanks so much to David, Giulio, and Hedda Hassel Morch for organizing it.

3. As you might have noticed, I’ve created a new category on this blog: “Obviously I’m Not Defending Aaronson.”  This category—reserved for posts that caused at least a hundred people to hate me—refers to a peculiar phrase I encountered over and over, in the social media threads denouncing me as a horrible person.  The phrase tends to occur in passages like: “look, obviously I’m not defending Aaronson, but it’s worth pointing out that, if you carefully reread everything he wrote, he never actually said that war orphans should be roasted alive and then eaten for fun.  That’s just something we all know that a clueless, arrogant nerd like him would think.”

4. Right now I’m at the “ThinkQ” conference at IBM in Yorktown Heights.  Here are the PowerPoint slides from my talk yesterday, entitled “The Largest Possible Quantum Speedups.”  Regular readers of this blog will find a lot that’s old and a little that’s new.

Follow