D-wave In D-news

Lots of blogging and press picking up on D-wave and Orion so I thought I’d collect a few here. The offical press release is here. I’d love to hear from anyone who has attended.
Scott Aaronson called me a Chinese restraraunt placemat in his The Orion Quantum Computer Anti-Hype FAQ (Update (3:17pm 2/13/07): Scott’s post now contains an update by the great Lawrence Ip, who now works for Google.) Doug Natelson, who gave an excellent talk here at UW a few weeks ago, poses three questions about the D-wave demo. Peter Rhode is every bit the skeptic and beats out Doug Natelson with four points. Ars-technica’s Chris Lee takes a shot at explaining adiabatic quantum computation and uses the word deathmatch here. You can find a bad quantum computing joke at the end of this blog post. I find this post amusing, if for nothing more than bringing politics into quantum computing. Coherence* remains the prettiest quantum computing website and has a choice Seth Lloyd comment “I’ll be a bit of a skeptic until I see what they have done. I’m happy these guys are doing it. But the proof of the pudding is in the eating.”
More mainstreamish media produces some truely incredible hype. One of my favorites is at physorg.com where we find the title a “New supercomputer to be unveiled” along with the choice gibberish “A Canadian firm is claiming to have taken a quantum leap in technology by producing a computer that can perform 64,000 calculations at once.” I flipped a coin 16 times today, can I get some venture capital? 🙂 Personally I like Gizmodo’s title: “D-Wave Quantum Computer to Span Multiple Universes Next Tuesday?” They also use the word sugerdaddy. If you want more reasons to be angry about hype or at bad journalism, go over to a wired gadget blog where you’ll find

There are certain classes of problems that can’t be solved with digital computers,” said Herb Martin, the firm’s CEO, over a decidedly-noisy digital cell phone. “Digital computers are good at running programs; quantum computers are good at handling massive sets of variables.”

Turing is certainly turning in his grave over that first sentence and, since Peter Shor is alive and well, I wonder if he is spinning today?
And don’t even get me started on this EETimes article. Choice:

Nondeterministic polynomial (NP) problems are the most difficult to solve on conventional computers because each variable adds yet another dimension to its possible solutions.

No, no, no! So many no’s I can’t even write it down. First of all NP problems include problems in P, so they definitely aren’t the most difficult to solve on a conventional computer. Second, the essentence of NP-complete problems is NOT just that you have an exponential search space. You’d think a Electrical Engineering rag would have taken some computer science courses? Then, of course EETimes only digs their grave deeper:

Quantum computers, on the other hand, can evaluate all possible solutions simultaneously and find the optimal solution, often in just a few clock cycles, thereby not only vastly speeding up the time taken to find the solution but also finding the most optimal result.

Okay, at that point I’ll admit I had to stop reading cus my brain was about to explode.
Oh, and whatever you do, don’t search for “first quantum computer” if you’ve ever performed a quantum computing experiment (that includes a lot of MIT Physics majors? Ack, is NMR quantum computation really quantum computation?) You might get a little miffed at all the years you spent in grad school doing what you thought were small quantum computer experiments.

13 Replies to “D-wave In D-news”

  1. I’m not sure the demonstration has actually happened, since the press release was already on the site many hours ago, when it was still night time at California!

  2. Some great quotes from this article:
    Quantum-computer technology can solve what is known as “NP-complete” problems.
    Quantum technology delivers precise answers to problems that can only be answered today in general terms. This creates a new and much broader dimension of computer applications,” Martin said.
    In general terms? What does this mean? And alledgedly it was said by Dwave’s CEO…

  3. yeah i believe the presentation is happening at like 3:30 pm today PST.
    Thanks again for putting all these article in one spot. I’ll have to click through em all this weekend.
    I’ll be attending the presentation on Thursday in Vancouver. Are you gonna drive up for it? If not I believe Geordie said they would be posting a video of the Mountain View event on their new website. I think he said it would be going up Thurday but don’t quote me on that.

  4. I was at the live event.
    The specific details of quantum coherence was addressed in a question that I asked and was answered by Geordie Rose. He indicated that they are submitting a paper for peer review for tests that they have done which provide evidence of the quantum effects that they are using.
    I do not think that paper will be the final word on that issue. However, they plan to get to 32 qubits with new IO by the end of the year and then 512 qubits Q1 2008. Clearly if they can tap into quantum effects for the 512 qubits then they will have a clearly measurable/apparent speed advantage. So they will build it. If it works then they will have an advantage and can sell things.

  5. While I don’t think you are in any way behaving like a placemat (or a quotation from a placemat?) in a chinese restaurant (doesn’t he really mean fortune from a chinatown fortune cookie?), I do rely on the QP for my BS filtering, so please don’t sugar-coat it: It’s your duty (as a scientist) to tell it like it is. I think I have to assume this D-wave stuff is crap until told otherwise, don’t I?

  6. And don’t even get me started on this EETimes article.
    Ah, that brings back memories. About twenty years ago,I saw some of my own work described in that publication.
    I didn’t even know it was possible to make 12 false factual assertions in a 10 word sentence, but they managed it.

Leave a Reply

Your email address will not be published. Required fields are marked *