However, I assume their creation operators are top-notch.
Scott Aaronson points out an overly-excited press release from NEC, which claims: “NEC, JST and RIKEN successfully demonstrate world’s first controllably coupled qubits”. This was indeed an exciting development when we published it five months ago. At best NEC has the world’s fourth controllably coupled qubits.
That said, the stupidity seems to be limited to the press release, and the paper actually looks pretty interesting, apparently with time domain results that no one else has shown. (I haven’t been on the campus network today so I haven’t had a chance to read more than the abstract.)
I thought about posting last night but this was pre-empted by the fact that the slides for my talk were unfinished (and also the Clarke group dinner). First I want to register a complaint:
This is how physicists (or maybe everybody) fill seating at conferences. The first people to arrive take the seats on the outside of the rows, and then fill in to the middle. This is really annoying when arriving in the middle of the session and having to climb over a bunch of people to get into the one empty seat. I am aware that this is a really lame complaint, but please, fill from the middle!
Now that I’ve got that out of my system: the last couple days were a blur of superconducting qubit talks. There’s a lot going on in this field, and most groups had three or four (10-minute) talks in a row to have enough time to explain all their results. One experiment I thought was very neat was this one from Terry Orlando’s group at MIT. In flux qubits like the ones we study, one can measure the temperature by sweeping the flux bias across the degeneracy point and measuring the population of the qubit states. Higher temperatures will give wider curves, as energies further away from the degeneracy point are more likely to be populated by thermal activation. When we measure this on our qubits we usually get something like 150 mK, mysteriously somewhat higher than the fridge temperature (roughly 50 mK).
What the Orlando group did was to apply an analog of laser cooling (as in atomic physics) to their qubit, using a microwave pulse to induce transitions that ultimately cool the system. As a result they were able to see these temperatures (as measured from the widfh of the qubit step) reduced by a factor of 100, from 300 mK to 3 mK. It was pretty impressive; I’m not sure how important it is for quantum computing or whether it’s something we should be doing with our qubits, but it’s a nice application of techniques from another field.
This morning I gave my talk, which was helpfully introduced by Frank Wilhelm’s talk immediately prior, in which he said something like “the really important development for scalability is what Travis Hime will talk about next”. So the pressure was on, but I think I did ok. After this was… more qubit talks, but I was mostly decompressing after finishing mine and didn’t pay as much attention as usual.
Tomorrow I go to see talks by other Clarke group members, including John himself. And then, an evening flight back to Berkeley.
Actually I spent much of today working on my talk instead of going to sessions. The superconducting qubit sessions start tomorrow morning and basically run continuously until Thursday evening. I did go to some talks in the afternoon, though, mostly in D2: Ion Traps for Scalable Quantum Computation. (In some sense this is our competition.)
Ike Chuang, who is a big name in this field, gave the first talk, which laid out the challenges in making a practical quantum computer with ion traps. Most of this dealt with error correction; according to Shannon’s theorem (or maybe a quantum information version thereof) it should be possible to build an error-free quantum computer out of qubits that do make occasional errors, as long as the failure rate is below some threshold. Unfortunately in some cases they’ve looked at this requires a prohibitively large number of operations, as many as 1020. One can try to implement various error-correcting codes, such as Shor’s or Steane’s, but certain operations that are needed for a universal quantum computer don’t work within these codes. And in fact Chuang et al. have shown that there is no stabilizer code that allows a universal set of operations to be performed within the code—one has to decode first before performing at least one of the operations.
The other talks in the session were less abstract, and thus harder to understand (since I’m not terribly familiar with this architecture). The talk by Slusher described a proposal for a VLSI-based scalable ion-trap based quantum computer, which seemed impressive, except I’m pretty sure this is the one Chuang mentioned that would require 440 watts of laser power to operate.
I skipped out on the last talk to go to D8: Superconductivity: STM of Cuprates and see what the group I worked in as an undergrad was up to. However, I haven’t thought about STM of cuprates for a while now and only had the faintest idea what they were talking about.
A tempting alternative for the end of the day was Session D33: Focus Session: Quantum Foundations II. It starts out as a perfectly normal session, but somewhere around 4:30 becomes the dumping ground for crackpots. For example:
D33.00014 : Do Particles have Barcodes?
If an elementary particle shown in Fig 2 of gr-qc/0507130 has an UNSTABLE quantum connection to the rest of the universe calibrated by nature in terms of Planck times, as also proposed in my separate MAR07 abstract, there exists a possibility that each particle has a barcode of its own. Instability implies varying periods of connections and disconnections of particles to the universe, which would be equivalent to the varying widths of white and black strips of commercial barcodes. Considering the high order of magnitude of Planck times in a second, each particle and the universe generated by its radiations may have their unique birth times registered in their barcodes. My quest for the cause of consciousness, in MAR06 abstracts, as an additional implication of physics/0210040, leads to the inquiry if these unique parallel universes are like the ones that give rise to consciousness as proposed by some physicists. With all due respect, the attempts to explain TOE of inert matter may not be attempts to explain one step to climb up on a stairway at a time. They may be attempts to explain only half a step at a time to on a stairway made with only integer number of steps. The search for TOE assumes such a theory exists. Mathematics has no barrels to fire bullets that can shoot down a non-existent bird. A Hamiltonian knows no consciousness, a missing ingredient of biology made of particles or vice versa, and of realistic TOE.
The talk after that one describes a theory of Atonic Physics [sic], which sounds like an outtake from Monty Python’s bookstore sketch.
Terence Tao explains quantum mechanics by analogizing to video games (particularly Tomb Raider):
Now, how does the situation look from Lara’s point of view? At the save point, Lara’s reality diverges into a superposition of two non-interacting paths, one in which she dies in the boulder puzzle, and one in which she lives. (Yes, just like that cat.) Her future becomes indeterministic. If she had consulted with an infinitely prescient oracle before reaching the save point as to whether she would survive the boulder puzzle, the only truthful answer this oracle could give is “50% yes, and 50% no”.
This simple example shows that the internal game universe can become indeterministic, even though the external one might be utterly deterministic. However, this example does not fully capture the weirdness of quantum mechanics…
He goes on to make some macabre modifications to the game mechanics in order to improve the analogy, bringing in interference and entanglement. It’s an entertaining post, but it gets truly ridiculous in the comments where he devises a Tomb Raider level to test Bell’s Inequality.
There’s been some buzz lately about D-Wave’s sixteen-qubit quantum computer that they’re planning to demonstrate tomorrow. Instead of writing a post on this I’m just going to link to (and endorse) Scott Aaronson’s post on the subject. There’s a lot of skepticism about D-Wave in the community.
As some of you know, we recently had a paper accepted to Science. The paper appears in the latest issue, and is now available online.
I will try to post something in the next few days that explains these results for the non-physicists in the audience. In the meantime, there’s this post from March about these experiments (from before we had the major findings), and here’s the abstract:
Solid-State Qubits with Current-Controlled Coupling
T. Hime, P. A. Reichardt, B. L. T. Plourde, T. L. Robertson, C.-E. Wu, A. V. Ustinov, John Clarke
The ability to switch the coupling between quantum bits (qubits) on and off is essential for implementing many quantum-computing algorithms. We demonstrated such control with two flux qubits coupled together through their mutual inductances and through the dc superconducting quantum interference device (SQUID) that reads out their magnetic flux states. A bias current applied to the SQUID in the zero-voltage state induced a change in the dynamic inductance, reducing the coupling energy controllably to zero and reversing its sign.
Here’s the latest publication on Clarke group qubit research, which appeared in Physical Review B at the end of May. Normally I give a non-technical explanation in these posts, but this paper is entirely devoted to working out gory technical details. It essentially goes through how to calculate a priori the properties of the flux qubits that I’ve written about previously. This calculation had been done for “small” qubit loops—small being defined in terms of the loop inductance but corresponding to a few microns on a side—our qubits are much larger than this (100 microns) and so we needed to figure out the more general solution.
The vast majority of the work in this paper was done by T. L. Robertson; my primary contribution was checking the math and the Mathematica code.
Quantum theory of three-junction flux qubit with non-negligible loop inductance: Towards scalability
T. L. Robertson, B. L. T. Plourde, P. A. Reichardt, T. Hime, C.-E. Wu, and John Clarke
Phys. Rev. B 73, 174526 (2006)
The three-junction flux qubit (quantum bit) consists of three Josephson junctions connected in series on a superconducting loop. We present a numerical treatment of this device for the general case in which the ratio betaQ of the geometrical inductance of the loop to the kinetic inductance of the Josephson junctions is not necessarily negligible. Relatively large geometric inductances allow the flux through each qubit to be controlled independently with on-chip bias lines, an essential consideration for scalability. We derive the three-dimensional potential in terms of the macroscopic degrees of freedom, and include the possible effects of asymmetry among the junctions and of stray capacitance associated with them. To find solutions of the Hamiltonian, we use basis functions consisting of the product of two plane wave states and a harmonic oscillator eigenfunction to compute the energy levels and eigenfunctions of the qubit numerically. We present calculated energy levels for the relevant range of betaQ. As betaQ is increased beyond 0.5, the tunnel splitting between the ground and first excited states decreases rapidly, and the device becomes progressively less useful as a qubit.
Via Mason, some guys at Caltech have set up a quantum information wiki intended for the research community. I added a page for myself, a stub page for the Clarke group, and updated their list of blogs to include this page and Mixed States. At the moment there’s not much there from the solid state angle, so I may be back to contribute a bit more.
Via Pharyngula, these tryouts for Stan Lee’s new superhero reality show remind me of nothing so much as the hero recruitment drive in Mystery Men. Perhaps I could use my quantum coherence research to develop a superhero persona, but my powers would only work if no one observes them. (Maybe this is just a secret identity requirement.) However, the field is probably rife with potential supervillainy.