Caption | Figure | Figure |
---|---|---|
HLT Tracking efficiency as a function of eta (left) and transverse momentum (right) for 1000 events ttbar simulated for the Phase-I detector with superimposed pileup collisions of 50 for two different algorithms. GPU Pixel tracking refers to the Cellular Automaton quadruplets pixel seeding algorithm executed on Graphics Processing Units. Triplet propagation algorithm refers to quadruplet seeds built by propagating triplets to the fourth Pixel layer. The ratio between the GPU Pixel tracking and Triplet propagation tracking efficiencies is also shown. No cut in transverse momentum was used to plot the efficiency as a function of eta. The GPU Pixel Tracking efficiency was optimized for transverse momentum higher than 1 GeV. The asymmetry in eta is caused by an inaccurate description of the Phase I Pixel detector's geometry.[Get pdf version left], [Get pdf version right] Contact: Felice Pantaleo |
Caption | Figure | Figure |
---|---|---|
Fraction of reconstructed tracks that are not matched to simulated charged particles as a function of eta (left) and transverse momentum (right) for 1000 events ttbar simulated for the Phase-I detector with superimposed pileup collisions of 50 for two different algorithms at HLT. GPU Pixel tracking refers to the Cellular Automaton quadruplets pixel seeding algorithm executed on Graphics Processing Units. Triplet propagation algorithm refers to quadruplet seeds built by propagating triplets to the fourth Pixel layer. The ratio between the GPU Pixel tracking and Triplet propagation tracking fake rate is also shown. No cut in transverse momentum was used to plot the fake rate as a function of eta. [Get pdf version left], [Get pdf version right] Contact: Felice Pantaleo |
Caption | Figure |
---|---|
HLT menu timing: Processing time distribution of the HLT menu /frozen/2016/25ns15e33/v4.0/HLT/V1, which correponds to one of the HLT menu used in the 2016 data taking operations. The timing was measured using events from run 279975, in lumi sections from 76 to 86, with an average instantaneous luminosity of 13.2 cm-2s-1 corresponding to a pileup of 42.5. The data was collected using the L1T selection, without any bias from the HLT. The processing time was measured on an "Ivy Bridge" Intel Xeon E5-2650v2 processor, running one single job. Last bin shows the overflow as well. [Get version without text], [Get version without grid] Contact: Mia Tosi |
HLT menu timing vs instantaneous luminosity: Processing time distribution of the HLT menu as function of the uncalibrated instantaneous luminosity, measured during data taking in fill 5393. Contact: Mia Tosi |