Learning?

Yesterday I spoke about almost everything except for what I have learnt at this summer school, so today I thought I would mention some of the things which I have learnt / heard over the past few days:

We had Trigger & DAQ talks given by the head of the CMS trigger, and it was quite interesting to hear what they were doing. I like their two level system, which seems more flexible than the ATLAS one, and also how they are going to add storage, so they can save events underground before sending them out to offline computing since the machine won't be running for long periods of time at the start. Their electron trigger is quite a bit more complicated than ATLAS though, since their calorimeters are inside their solenoid, meaning the energy deposits are asymetric and have strange shapes. They are also implementing a pseudo-RoI approach to reduce the amount of data they need to unpack, which I didn't realise they were doing.

"There are many physicists in the world... but not so many good ones" - Guido Atarelli, of the DLAP QCD evolution equation. His lectures and discussion sessions have been fairly interesting due to his way of putting things. He has been giving the theoretical introduction for the school, lying the basics for the more detailed lectures on the various areas. He has also mentioned a few discrepancies between experiment and theory which I didn't know about before, though the differences are only about 3 sigma.

I have also heard a fair bit about QCD, but I don't know how much I have acutlaly learnt from those lectures, since most the of the content is new to me, and I find it hard to follow theory from lectures without the follow up exercises and reading.

We also had a Computing talk by a person high up in CMS computing, in which I learnt that CMS doesn't have the same Tier 2 cloud structure around the Tier 1s. Theoretically, each of their Tier 2s should be connected to each of their Tier 1s, though in reality, this isn't true, but it only matters for raw data, since each of the Tier 1s should have a full copy of all the analysis level data, though this won't be true for first data.


This is the FermiLab logo. Weird looking isn't it? Today I learnt that it's a schematic of a quadrupole superimposed on a dipole magnet as the Tevatron here was the first synchrotron to use quadrupole magnets in a FODO lattice to focus the beam instead of extended aperture dipoles.

I got the answer for a question I've been pondering for a while, which is, how the proton beams in the LHC will pass each other in the single beam pipes in the experiments when not colliding. And it's complicated trajectory which allows them to pass each other without colliding, made by not turning on the focusing magnets. I guess I should have realised that this was possible, since other machines only have one beam pipe all the way around, like the Tevatron, and they can manipulate the beams there to only collide when they want it.