Biggest quantum cluster does not compute ... yet
10,000 entangled modes can be 'programmed' just by measuring them
Video It's currently the biggest thing in the smallest world: researchers from the ANU, Sydney University and Tokyo have created the world's largest entangled state to date.
Demonstrating a commendable why-do-it-by-halves approach, the researchers have done rather better than a “quantum leap” here: they say they've created an entangled resource, called a “cluster state”, that contains more than 10,000 entangled modes.
That's more than three orders of magnitude better than the previous record (set in 2012) for a cluster state, which had eight entangled photons.
As the researchers explain in the abstract of their paper in Nature Photonics, larger cluster states enable larger computations. With 10,000 entangled modes, there's an awful lot of superposition that can be exploited by a suitable quantum computing algorithm – eventually, which we'll get to later.
To get a grip on what's going on, The Register spoke to Dr Nicolas Menicucci at the University of Sydney, who proposed the scheme the project adopted and is a co-author of the paper.
The simplest illustration of the process is in the video below by Seiji Armstrong, a visiting student from ANU who worked on the experiment in Tokyo.
To create the 10,000-mode cluster state used a relatively simple hardware setup, he explained: two light beams, both already in what's called a “squeeze state” are run through a beam splitter. This creates entangled states between individual “chunks” of the beam.
That, however, isn't 10,000 modes. To get that, one of the beams is given a time delay (by running it through a longer piece of fibre – the hardware setup is that simple!). By running the “future beam” (the delayed beam) and the current beam (no delay) through a second beam splitter, a past-future interaction is created that sets up the long string of modes.
The cluster state isn't yet ready to perform quantum computations. However, Dr Menicucci told The Register that much of the theoretical groundwork for computation already exists.
“The basic protocols are there, but they don't cope well with noise. The crystals [which put the light into its squeeze state] determine how noisy the computation will be, and they can never be perfect.
“So we need protocols to tolerate the noise that's inherent in these systems.”
As it now stands, however, the 10,000-mode cluster state isn't yet suitable for quantum computation: without further work, it can only process a single qubit (quantum bit) worth of quantum information, Dr Menicucci said.
What's needed is to give what is currently a one-dimensional output a two-dimensional shape. To do that, his proposal uses a second copy of the same experiment (in addition to the first) with a longer fibre delay, plus extra beam splitters to weave the final state together at the end.
As a visual metaphor, Dr Menicucci offered thread wound onto a spool: the same thread touches itself at a host of different points. With that metaphor in mind, he explained that a two-dimensional surface offers more places for different qubits to interact with each other.
Each of those qubits provides the mechanism by which information can be encoded onto the cluster state.
And how would the as-yet theoretical quantum computer do its work?
Here's where the quantum universe slips gently from El Reg's mental grasp.
“What you do next is that you make measurements on the state,” Dr Menicucci explained. “The choice of measurement determines what computation occurs. In what's known as 'back action', the act of measurement changes the system you've just measured.”
The choice of the sequence of different measurements the observer makes on the system “means that you manipulated the information that's encoded into the state”, he said. ®