Quantum Bayesian Networks

September 30, 2010

Enough QC Hardware Approaches To Titillate, Confuse and Annoy

Filed under: Uncategorized — rrtucci @ 8:28 pm

Quantum Computing has appeared frequently in the news this year. And deservedly so, because it is making steady progress. On Feb.17, I posted an entry titled “Experimental Quantum Computing, State of the Art”, where I discussed an excellent review article of QC hardware written by Franco Nori and his team. Now, 8 months later, that review article is already seriously out of date.

I just counted how many stories NextBigFuture has had since the beginning of the year under the subject of “quantum computer”:

http://nextbigfuture.com/search/label/quantum%20computer

34 stories from Jan 1 to Oct 1. All in all, a pretty exciting year if you ask me. Many of those stories are not earth-shaking events, but some are. For instance, measuring the spin state of a single electron (the electron belongs to a donor Phosphorus atom implanted in a Silicon crystal). Wow!

The public might be understandably confused. What hardware approach will win the race for a scalable quantum computer. Is it going to be ???

  1. an ion trap quantum computer (NIST),
  2. or a superconductive device that implements an adiabatic QC à la D-Wave,
  3. or a superconductive device that implements the gate model (Yale and UCSB)
  4. or an array of donors implanted in Silicon by the Australians and Finns,
  5. or NV color centers in diamond,
  6. or a photonic QC by the Brits,
  7. or Rydberg atoms
  8. or an anyonic QC

I probably left out a few.

Makes me glad I specialize in writing quantum computing software, which is somewhat platform independent, so I don’t have to guess which of all these wonderful approaches will win out.

September 17, 2010

Early Failures in the History of Quantum Sampling

Filed under: Uncategorized — rrtucci @ 5:35 pm

Langley Aerodrome experiment (October 7, 1903)

On Dec 1999, Grover published a paper entitled “Rapid Sampling Through Quantum Computing” arXiv:quant-ph/9912001 . In it, Grover gives a method for sampling a probability distribution P(x) using a quantum computer. His method does this “in O(sqrt(N)) steps. A classical algorithm would need O(N) steps.” In a nutshell, Grover’s sampling method applies the original Grover’s “search” algorithm to N_B + 1 qubits, with a starting state:

|s'\rangle = |0^{N_B}\rangle\otimes|0\rangle

and a target state

|t\rangle = \frac{1}{\sqrt{2^{N_B}}}\sum_{x\in\{0,1\}^{N_B}}|x\rangle\otimes\left(\sqrt{P(x)}|0\rangle + \sqrt{1 - P(x)}|1\rangle\right)

Once this starting state is steered to this target state, one measures the target state to sample P(x). A serious problem with this method is that such a target state is useless for sampling P(x), because it has a vanishingly small amplitude of size O(1/\sqrt{2^{N_B}}) for all x. Quibbs does not suffer from this defect because its target state is

|t\rangle = \sum_{x\in\{0,1\}^{N_B}}\sqrt{P(x)}|x\rangle\otimes|0\rangle

Thus, Quibbs’ target state has a finite amplitude at those x for which P(x) is finite.

ACHTUNG: (added on Oct. 11) This blog post is a scurrilous lie. See retraction here.

September 9, 2010

Quantum Computing Jobs

Filed under: Uncategorized — rrtucci @ 3:48 pm

Most of the people who read this blog are interested in quantum computers, and some of them are even interested in pursuing careers in quantum computing. So let me try to describe here the current QC job market.

Prof. Todd Brun, a very considerate chap, maintains a fairly complete list of available jobs in quantum information processing (QIP)(quantum computing is considered a subset of QIP.) Brun’s list currently contains

  • 10 ads for post doctoral positions
  • 0 ads for faculty positions or permanent positions at government labs.
  • 1 ad for an industrial job, albeit in the bogus field of quantum cryptography
  • 0 ads for co-ops or internships. No ads for this one is particularly disturbing to me.

As you can see, QC faculty position are extremely rare, a vanishing species. I don’t know the exact numbers but my guess is that about only 1 out of every 20 QC postdocs ultimately gets a tenured faculty position. And postdocs can look forward to 2 to 4 postdoctoral jobs (each lasting 2 years). Of those fortunate enough to finally get an assistant professorship at a university, maybe half(?) will be denied tenure.

Many talented people interested in quantum computing, people who could greatly advance the QC field, choose not to enter it, or they leave it prematurely. Who can blame them. The QC academic career path is so risky and taxing, and, unfortunately, there is no other QC game in town. There are currently no QC career paths besides the academic one. It would be nice if industry were providing alternative careers in QC, but this is not happening at present. Industrial QC jobs are currently more rare than the dodo bird.

In my opinion, American industry could be and should be creating at least a few QC jobs right now, but they are failing to do so.

  • If you look at who is currently making important contributions to experimental quantum computing, you won’t find the names of any big hardware companies like IBM, Intel and HP, or of any startups. Instead, you’ll find only names of universities and NIST. (D-Wave is the only shining exception to this rule).

  • You would think that big software companies like Google and Microsoft or maybe startups would be busy producing software for QCs, but no. (Hartmut Neven et al, from Google, wrote a bit of software for D-Wave’s machine circa Dec 2009, but I don’t know if that work is ongoing, and Neven must be too busy as head of Google Goggles to devote much time to QCs). (Microsoft has a small group of people (StationQ) working on QCs, but they are mostly abstract mathematicians trying to do experimental physics. Sort of like cats trying to act like dogs :). I don’t think they are writing any QC software).

American industry is conducting almost no QC research. They want government-funded projects to do all the research and invention. After the government does all the heavy lifting, they want to reap the rewards themselves. It’s an example of industry expecting our government to fully fund activities which are not the responsibility of a government. Such attitudes on the part of industry are very unhealthy for our country, because they slow technological progress and inflate government debt. Such attitudes are also a bad business decision which delays or diminishes the profit for businesses.

September 6, 2010

(Guest Post) Hype, Hope and Building the Quantum Community

Filed under: Uncategorized — rrtucci @ 3:12 pm

(Note from rrtucci: My friend Matt Swayne wrote the following piece about Suzanne’s talk. I totally agree with him. Suzanne Gildert rocks)

I had the chance to attend Suzanne Gildert’s virtual seminar on quantum computing Saturday.

Dr. Gildert is a physicist who works for D-Wave Systems, not a pioneer of quantum computing, but the pioneer in quantum computing. She’s also a blogger at Physics and Cake .

Besides serving as an overview of quantum computing systems and technology, the session was aimed at delineating between the hope and hype of quantum computing. Just like any technological revolution, quantum computing seems to divide people into two distinct camps: Luddites and Utopians. Half think that quantum computing will never happen; half think that quantum computing will automatically create a technological utopia.

Both sides are probably wrong.

After attending the seminar, I believe that we’re within a few years of a new era in quantum computing. In Winston Churchill’s words, it won’t signal the end of the quantum computing revolution. It might not even be the beginning. But it may be the end of the beginning.

I also think that Suzanne did a good job of setting expectations for what this revolution will actually mean. I don’t think QCs will change society overnight, or lead to AGI out of the gate. But it does give us an impressive tool to work with.

Following Suzanne’s presentation there was an extended discussion period about not just QC, but other technologies and the philosophical implementations that surround the notion of quantum computing–the multiverse, many worlds theory, etc. All interesting stuff.

The real takeaway for me was something that Suzanne mentioned during the chat. Hopefully more private groups and companies, who won’t be as confined in their research as government and higher ed. researchers often are, will incubate around quantum computing. That’s absolutely necessary to create a QC industry. One thing is certain, I can tell just by the level of conversation after the presentation that there is enough brain power already in the sector to build this industry.

Regrettably, Suzanne mentioned nothing about cake during the session. Perhaps a follow-up is in order.

September 3, 2010

Double Treat: The Latest in Adiabatic Quantum Computing and Teleconferencing

Filed under: Uncategorized — rrtucci @ 4:07 pm

Dr. Suzanne Gildert will give a talk in Teleplace on “Building large-scale quantum computers: Fundamentals, technology and applications” on September 4, 2010, at 10am PST (1pm EST, 6pm UK, 7pm CET). For more info about the talk, see this blogpost by her.

Suzanne is a very clear and well-informed explainer. She works as an experimental physicist at D-Wave, a company based in Vancouver, Canada, that does leading edge research in adiabatic quantum computation.

Telespace is an innovative teleconferencing software with an interface a bit like Second Life.

Blog at WordPress.com.

%d bloggers like this: