The Top 500 list of the most powerful supercomputers has been disclosed during Supercomputing '12

The most surprising here is the Blue Waters system for which no hpl numbers were submitted.
Bill Kramer explained the reason why it was decided not to submit anything, now or ever (see here).
Bill is very right when explaining why hpl makes very little sense today.
However, i think he misses an important point : submitting hpl results simply means the system was able to run it uninterrupted and without suffering any slowness. This might looks trivial from a rhetoric point of view ... but back to the real life, having a big system (e.g. thousands of nodes) running at full speed without any single failure for tens of hours is quite a bit of a challenge and hence a (small but important) sign of the stability of such a supercomputer.