News from the LHC: the integrated proton-proton luminosity at 7 TeV centre-of-mass energy has generously passed the mark of 40 inverse picobarns yesterday. The CMS experiment alone has integrated over 42 inverse picobarns, as shown in the graph below (the blue curve shows the data collected by CMS, the red one the data produced by the LHC).

The closeness of the two curves demonstrates how the CMS experiment is operating at full functionality, collecting data with high efficiency. Actually, it should be mentioned that a 100% data collection efficiency at a hadron collider is not the best of worlds: 100% efficiency suggests that the trigger system might be operating a too tight selection, incurring in no dead time as it attempts to sort out which ones among the collisions is worth collecting and saving to disk. Of course, until the accelerator runs with all the proton bunches in their place along the beam, the trigger system is not strained to its maximum; right now, only about 400 of the 2808 possible bunches are filled with protons, and their distribution along the beam has an impact in the data acquisition chain. In general, though, some dead time is an excellent trade-off ground with the accuracy with which interesting collisions are sorted out. For instance the experiment CDF at the Fermilab Tevatron, which has a 25 years old experience in triggering hadronic collisions, typically runs with a 5-10% dead time caused by the trigger system.