“If a tree falls in a forest, and no one is around to hear it, does it make a sound?” is a well known metaphysical toy question. What is less known is its business corollary: If a start-up makes a splash, but nobody pays attention, did it actually ever really emerge from stealth mode?
It is rather welcome news that the latest Nobel prizes in physics bring some attention to quantum computing, but as always, these machines are presented as a future possibility rather than a commercial reality.
Yet, D-Wave already ships the first commercial quantum computing device, and hardly anybody seems to know. Sometimes this ignorance appears to be almost willful. The company just doesn’t seem to be able to generate ubiquitous mainstream media awareness.
While their device is not a universal quantum computer, it nevertheless can facilitate an extremely useful and versatile quantum optimization, with plenty of real life application usage. The somewhat arcane Ramsey number research already demonstrated the power of the Rainier chip and a recent paper published in Nature shows how NP complex protein folding scenarios can be calculated on their hardware. An interesting wrinkle to this latest research is that the device doesn’t find the optimal solution as reliably as in the Ramsey number case, but that the other solutions are also valid folding modes of the examined protein and provide additional insight.
The importance of this kind of research can hardly be overstated, so why is this company not mentioned every time somebody writes about quantum computing?
Is it a case of crying wolf too many times? Is there a perception that the company may have over-promised and under-delivered in the past? Is it a lingering after-effect of the Scott Aaronson ham sandwich controversy?
Your guess is as good as mine. If you have any thoughts or insights on this please share them in the comments section.
Good question! I suspect it is simply inertia. The D-Wave device seems to me to be best understood as a quantum analog computer. This creates a difficulty for them since folks think computing = calculation = digital architecture.
Simple thought chains like these become second nature, even to well-trained and thoughtful people. When you have a radical innovation folks can take a long time to realize the sense of it. Really radical innovations change more than one assumption.
D-Wave is an incredibly radical innovation since it is both quantum and analog and it requires a whole different programming mind-set (generating functions).
I think this may have something to do with it.
When you take more than one radical step it is terribly hard for conventional minds to comprehend. It just will seem wrong to them. Alien even.
I think digital computation will come under broad threat in coming years. The problem with digital is that it is intrinsically “jumpy” in an obvious way. This is exceptionally bad thermodynamically as your device is far from adiabatic and isentropic. As a result, they do not scale well thermally.
This is why some conventional computing projects (classical) are now pioneering large scale parallel analog computation. This is a pure piece of engineering.
Once you understand the above points, it becomes clear that massively parallel analog computers are likely to have far superior space, energy and thermal performance.
It is a hunch of mine that people will “get” D-Wave once they “get” that analog computers are perfectly good computers.
They were the first computers and, in the final analysis, probably the best computers.
Once that happens some prankster at MIT will write a book about how the universe is really a giant brain and not a giant computer. The public will buy it in droves 🙂
It’ll be quite ironic if the computing industry is about to come full circle like this.
It is a little know fact (due to decades of being classified) that analog computing was instrumental in Brittain’s WW2 air defense. Combing radar tracking with analog computed flight path projections enabled the UK to neutralize the V1 (a jet powered pre-cursor of all modern cruise missiles).
Great picture Henning!
D-wave is doing fine for now. The latest Bezos et al cash infusion should allow them to prove their worth (or not). I do hope that other private companies attempting to build a gate-model QC sprout too…and soon. The horse race between the adiabatic and gate models promises to be very exciting
Thanks! The picture really tested my patience. The first time around, immediately after completing it, my Laptop crashed before I hit the save button. Fortunately I am getting faster at working with Gimp.
The picture turned out to have a bit of a Dr. Who feel to it but I think that’s entirely not a bad thing …
Would very much like to see a race of QC architectures but there’s no guarantee that the best technology always wins. In my experience marketing plays a huge role in the IT arena.
I think it’s mostly because they won’t tell anyone how their tech works or even what it really does. They say they’re doing “quantum” computation, but who knows? Of course all modern transistors work on quantum principles. Presumably they’re doing something more interesting than that, but until they publish papers everyone will have to guess. They’re trying to be a 19th century proprietary-tech company in a 21st-century open world. After their initial announcements last year (or was it two?) where they unveiled “something” but wouldn’t let anyone near it, or try it, or anything, I expect most people wrote them off as snake oil salesmen. It’s hard to counter that once the perception is formed.
That’s the problem. Perhaps (or perhaps not) they’ve proven they’ve built a special-purpose computer for solving one specific optimization problem. But its unclear if this is achieved using quantum entanglement — or just “state of the art” classical effects. Nor have they shown that even if they prove that they can solve quadratic unconstrained binary optimization problems embeddable on their chip (“embeddable” being the key issue), that such solutions don’t result from just really good engineering using basically classical effects. Look, what they are doing is ambitious and it would be fantastic if it all proves out, but until that can be shown, I think it’s better to remain hopeful but skeptical.
Hey Gary,
Just for the record … we’ve been extremely transparent about what our processor is, the theory behind it, and the justifications for calling it a quantum computer …. We’ve published a large number of papers on every aspect of our processor in peer reviewed journals (Nature, Phys Rev Letters, etc.) … all these papers are easily accessed on our website (or the archive) … not to mention that a impressive (and skeptical) team of both theoretical and experimental physicists at USC/ISI have been experimenting with and confirming our results as an independent party … Additionally, we’ve been running real world applications with Google (Image Recognition), Harvard (protein folding), etc. detailed descriptions of which are also available on-line
I think the reason for the disparity in perception between what D-Wave has achieved and how it’s represented in much of the press is a sociologic phenomenon common to all ambitious efforts that challenge mainstream/incumbent development efforts ….. first, such new efforts are often seen as a threat to the status quo efforts (in terms of funding, exposure etc.) and there’s both conscious or unconscious efforts to discount them … second, the researchers entrenched in their own paradigms and communities seldom take (or have) the time to research new, non-standard approaches and hence don’t understand them … often try to re-interpret the new approach in terms of formalisms they understand (like applying de-coherence time quality metrics in the gate model to the adiabatic model where’s it’s not at all relevant in the same way) causing mis-conceptions about the validity or potential of the new approach
It’s fascinating to me that researchers interested in quantum computing and knowing of D-Wave’s existence, it’s constant stream of technical results, it’s world class team of scientists/engineers (hired from some of the best traditional QC groups), the scale of the effort … don’t buy a plane ticket and visit … or at the very least read our papers, call our scientists and investigate more what we’re up to more scientifically. Everyone who’s visited and taken the time to understand what we’re doing has been impressed and excited by our progress and the future potential.
So, Gary, and anyone else who is interested … read our voluminous papers, attend our talks, or talk to our scientists, and you’ll get detailed descriptions of what are processor is, and does … I think you’ll be pleasantly surprised
Eric Ladizinsky
Co-Founder, Chief Scientist
D-Wave Systems
Hi Eric; thanks for taking the time to reply to a lowly commenter! I wonder what you think about the new Tech Review article (http://www.technologyreview.com/news/429429/the-cia-and-jeff-bezos-bet-on-quantum-computing/). It’s largely positive, but seems to have some of the same reservations I expressed above.
I think people just don’t completely understand WHAT a quantum computer is and how it can be useful to them (even from a commercial point of view). Many people tend to think that a quantum computer is a massively-parallel traditional computer, which couldn’t be further from the truth.
Henning – the picture is awesome!
As to why everyone is not jumping up and down about Dwave yet, Kingsley pointed out a good reason. Computation=digital mindset is geared up towards conventional, familiar ways of computing. AQC to most people is somewhat like a computer with only assembler language supplied. Ok, great. So what a regular offce user would do with it? When the urgent need for this link between great hardware and real application world has been realized, Dwave expanded its application group to particularly address this need – since we are the ones who understand the hardware best, we are the ones to create this link. Also what should be mentioned is that really only now the hardware broke the threshold of actually outperforming best heuristics tailored to the Dwave’s architecture. Guys at Lockheed understood the potential and agreed to work with us to develop the link to solve their problems, but it’s not quite there yet. Most customers would want out-of-the-box solutions for their problems, and this is the show-stopper. When e.g. Lockheed will announce that they are now efficiently solving their (real, commercially/strategically-important for them) problem of such and such with Dwave’s QC (which probably won’t happen due to the internals of their businesses being highly classified) – that is when everyone will wake up.
Elena, most certainly building business references is the key marketing task for any IT company. Lockheed is a feather in the cap, but you are right that the classified aspects of their business may make for difficult reference building. The military-industrial complex has deep pockets but I am looking forward to more generally applicable solutions such as supply-chain-, inventory-, offer- optimization etc.
Wished I had more time test driving your development environment. Very much like what your company put out – Python is one of my favorite languages 🙂
Elena, D-Wave’s computer is not analog. It is a digital computer. I do not know about any analog quantum computer at the moment.