The linked article is a little confusing. Here's a better description:
Pairs of quantum-mechanically entangled particles seem to know at once what is happening to each other. Experiments show that even if this signalling is not instantaneous, it must be really, really fast.
Ben - All of those are sitting behind a publisher paywall :-( Do you have a quick summary of how this goes beyond previous tests of the Bell inequality?
Suppose we try to explain quantum correlations by postulating that as soon as one member of an entangled pair of particles is measured, a very fast signal travels to its partner, updating its state. If the signal is instantaneous, then such a model is indistinguishable from quantum theory. But what if the signal is fast, but not infinitely fast?
Salart and collaborators establish a lower bound on the speed of any such signal (it must be > 10,000 times the speed of light). But we have to be careful about which reference frame is used to define the velocity. Natural choices, such as the rest frame of the experiment or the rest frame of the cosmic microwave background, were ruled out by an earlier experiment, also by Gisin's group [1]. The new experiment makes use of the rotation of the Earth to establish a lower bound that holds for any reference frame.
Pairs of quantum-mechanically entangled particles seem to know at once what is happening to each other. Experiments show that even if this signalling is not instantaneous, it must be really, really fast.
This is from Terry Rudolph's commentary:
http://www.nature.com/nature/journal/v454/n7206/full/454831a...
The Editor's summary is also good:
http://www.nature.com/nature/journal/v454/n7206/edsumm/e0808...