No, this is not an obituary for D-Wave.
But the reporting of the latest news connected to D-Wave just doesn’t sit well with me.
Ever tried to strike up a conversation about Ramsey numbers around the water cooler, or just before a business meeting started? No? I wouldn’t think so.
I don’t mean to denigrate the scientific feat of calculating Ramsey numbers on D-Wave’s machine, but the way this news is reported, is entirely science for science’s sake.
It puts D-Wave squarely into the ghetto of specialized scientific computation. Although, I am convinced that quantum computing will be fantastic for science, and having a physics background I am quite excited about this, I nevertheless strongly believe that this is not a big enough market for D-Wave.
It is one thing to point to the calculation of numbers that less than one out of ten CIOs will ever have heard of. It is another matter entirely not to milk this achievement for every drop of marketing value.
In all the news articles I perused, it is simply stated that calculating Ramsey numbers is notoriously difficult. What this exactly means is left to the reader’s imagination.
If your goal is to establish that you are making an entirely new type of super-computer then you need an actual comparison or benchmark. From Wikipedia we can learn the formula for how many graphs have to be searched to determine a Ramsey number.
For R(8,2) D-Wave’s machine required 270 milliseconds. This comes to more than 68,719 million search operations. For a conventional computer one graph search will take multiple operations – depending on the size of the graph. (The largest graph will be 8 nodes requiring about 1277 operations). Assuming the graph complexity grows with O(2n) I estimate about 800 operations on average.
Putting this together – assuming I calculated this correctly – the D-Wave machine performs at the equivalent of about 55 million MIPS. For comparison: This is more than what a cluster of 300 Intel i7 Hex core CPUs could deliver.
Certainly some serious computational clout. But why do I have to waste my spare time puzzling this out? At the time of writing I cannot find a press release about this on the company’s web site. Why? This needs to be translated into something that your average CIO can comprehend and then shouted from the rooftops.
D-Wave used to be good at performing marketing stunts and the company was harshly criticized for this from some academic quarters. Did these critics finally get under D-Wave’s skin?
…. I hope not.
Update: Courtesy of Geordie Rose from D-Wave (lifted from the comment section) here is a link to a very informative presentation on the Ramsey number paper. While you’re at it you may also want to check out his talk. That one definitely makes for better water cooler conversation material – less steeped in technicalities but with lots of apples and Netflix thrown in for good measure. Neat stuff.
Update 2: More videos from the same event now available on D-Wave’s blog.
My 2 cents:
I also think D-Wave should sell itself as an agnostic superconducting QC company. If gate level superconducting QCs ever become competitive with D-Wave’s adiabatic QCs (which hasn’t happened yet, and may never happen) D-Wave should assure its investors that they have the expertise in superconducting tech to jump on the gate-model band wagon too.
Gordie claimed that they know how to tweak their processor to make it universal equivalent:
He’s on record of expressing many times that he doesn’t think Shor, Grover etc. are commercially interesting enough to go there.
Found in the comment section here.
In this case they really hold a lot of trump cards and have every reason to rev up the marketing noise.
I was at the NASA conference on quantum computing last week and the guy (Frank Gaitan) who did the Ramsey numbers work had a great presentation about it. It was really very compelling. I believe the reason there is no press release etc. from dwave is that the work wasn’t done by dwave it was done by the NSA independently using a dwave system.
Interesting. Wouldn’t be surprised if there was a D-Wave system sale to the NSA that was kept under wraps.
To my knowledge there is no algorithm in the open that can quickly factorize primes with D-Wave’s machine but that doesn’t mean the NSA might not be trying.
No need to worry. Chinese cyberspies probably read everything except this blog.
Don’t shatter my illusions! I really counted on this Chinese audience. C’mon sprinkling in “NSA”, “Quantum Computing” and all that surely must trigger a tripwire? 🙂
quantum lurker said:
“I was at the NASA conference on quantum computing last week and the guy (Frank Gaitan) who did the Ramsey numbers work had a great presentation about it. It was really very compelling. I believe the reason there is no press release etc. from dwave is that the work wasn’t done by dwave it was done by the NSA independently using a dwave system.”
Hmm, interesting
The d-wave/ramsey numbers paper
http://arxiv.org/abs/1201.1842v1
lists Gaitan as being a member of LPS
http://www.lps.umd.edu/
which portrays itself as an academic institution. Or would it be more accurate to say that LPS is essentially a “front” for the NSA?
Note that the first author of the d-wave/ramsey numbers paper is a chinese expatriate
http://ca.linkedin.com/pub/zhengbing-bian/8/768/132
That must not sit too well with the NSA.
Another mysterious fact is that Geordie originally had a blog post on the Ramsey numbers paper. But for some mysterious reason he deleted it. It’s still cached in google though
http://webcache.googleusercontent.com/search?client=safari&rls=en&oe=UTF-8&hl=en&q=cache:zsB21rMRCpIJ:http://dwave.wordpress.com/2012/01/11/using-the-d-wave-one-to-calculate-ramsey-numbers/+d-wave+ramsey&ct=clnk
When I searched for any reaction from D-Wave on this news I also came across Geordie’s google cache fragment. At the time I didn’t make much of it – thought it was because they were preparing a proper press release. But there’s still none up. Bit odd, isn’t it.
Here is a link to Frank’s talk referred to above: http://connect.arc.nasa.gov/p1zygxics39/?launcher=false&fcsContent=true&pbMode=normal.
Mine is here: http://connect.arc.nasa.gov/p69sps0r4bc/?launcher=false&fcsContent=true&pbMode=normal.
Thanks for the links. Quite enjoyed your presentation. The image feature extraction reminds me a bit of a more generic wavelet decomposition (sans wavelets) while retaining the robustness of a Hopfield neural network trained to reconstruct images. Pretty neat stuff.
With regards to Frank’s talk – did he give more than one? The linked one deals with Spectral Gap Amplification but not the Ramsey number calculation?
whups wrong link, it’s here http://connect.arc.nasa.gov/p38z3xb8x5t/?launcher=false&fcsContent=true&pbMode=normal
Updated the blog post with the links provided by Geordie. Based on the presentation, I think we can put to rest the rumour that quantumlurker started with regards to the NSA having a D-Wave machine.
This will be very disappointing for my intended audience of Chinese cyberspies. Just for their benefit I suggest the possibility that the NSA outsourced its QC activities to Vancouver. As Canadians usually come across as mostly harmless this is obviously a very cunning ploy.
Updated the blog post with the links provided by Geordie. Based on the presentation, I think we can put to rest the rumour that quantumlurker started with regards to the NSA having a D-Wave machine.
This will be very disappointing for my intended audience of Chinese cyberspies. Just for their benefit I suggest the possibility that the NSA outsourced its QC activities to Vancouver. As Canadians usually come across as mostly harmless this is obviously a very cunning ploy.
“As Canadians usually come across as mostly harmless this is obviously a very cunning ploy.”
LOL. Rumors about the death of Nortel and RIM have been widely exaggerated.
The new rumor is that China has a quantum computer. Don’t underestimate your NSA readers.
“As Canadians usually come across as mostly harmless this is obviously a very cunning ploy.”
LOL. Rumors about the death of Nortel and RIM have been widely exaggerated.
The new rumor is that China has a quantum computer. Don’t underestimate your NSA readers.
I think the comparison to 300 CPUs is far fetched. Using a simulated classical annealer I have solved the same Ramsey number problem in less than a millisecond on a single CPU core.
My assumption was a brute force search of the entire parameter space as alternative (assuming that quantum annealing always finds the correct answer with extremely high probability). There are certainly plenty of heuristics to quickly find these rather simple Ramsey numbers.
Simulated classic annealing won’t do the trick for larger graphs (i.e getting stuck in local minima) and my expectation is that actual quantum annealing will fare better.
That being said, things will certainly only get really interesting with higher integrated chips that’ll allow to embed larger graphs to get to some none-trivial Ramsey numbers.
In this sense the result they got here resembles the “feats” of the research on gate based quantum computers (e.g. Shore’s algorithm factoring 21).
What makes D-Wave most interesting is that they seem to constantly double their chips integration density year over year, so hopefully we don’t have to wait too long before this experiment can be repeated for more meaningful Ramsey numbers.