land of freedom and democracy :)))))))))

Given that it is an "embarrassingly parallel application" (its distributed computing not parallel computing) things like CUDA, OpenCL or MPI are excluded.
You, first of all, must make clear to yourself if you need threads of executions or cores for processing.
Price wise i would recommend multiple dual socket AMD servers: best price per core and if your analysis is scaling linear with cores, it is the best choice.
For administering such nodes i would recommend rocksclusters.org (underlying job scheduler can be torque, condor or sge)

Scientific Linux is a clone of Red Hat Enterprise Linux (like CentOS). As such it follows closely the versions of RHEL and basically you cannot make difference between SL and RHEL.

More on the subject : for the data analysis one can use ROOT (root.cern.ch) which is de facto tool for data analysis and simulation at CERN. It is used from online data taking from detectors to offline processing (even distributed - PROOF) of data.

IAAPP and that part with black holes is pure fiction. A black hole is a singularity at level of space-time fabric. Quark Pluon Plasma is a state in which quarks and gluons are almost free because of high confined state. From what we know until _now_ one has nothing to do with the other.However, __IF__ they(black holes) will be produced, the evaporation time will be at level of ns. So .. the afirmation that: "one of the purposes of this huge facility is to generate small blackholes" is completly false. Better read at www.cern.ch and also http://aliceinfo.cern.ch/ (home page of ALICE experiment) which is an experiment dedicated to study of Quark Gluon Plasma