an update November 2013 Norman Gee Caveat The spreadsheet calculation has never agreed with the measured numbers Of course Thilos measurements are right and the spreadsheet is wrong Need to understand the ID: 487794
Download Presentation The PPT/PDF document "1 Phase I Latency Envelopes –" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
1
Phase I Latency Envelopes – an update
November 2013
Norman GeeSlide2
CaveatThe spreadsheet calculation has never agreed with the measured numbers
Of course Thilo’s measurements are right and the spreadsheet is wrongNeed to understand the differences – more on this in a momentThe update I did
for a recent L1Calo meeting showed only about 5BC of spare latency remainingI
recovered
some latency for Upgrade Week by better modelling of CP, JEP, L1Topo and CTPThese have multiple inputs and/or outputs at different timesE.g. CTP has electrical into CORE, electrical into Input modules, Optical into CORE, all with different latenciesThe spreadsheet version used here is a working version intended to give the latest estimate, it is not yet stable enough to be uploaded into EDMS
Oct-2013
Latency - N. Gee
2Slide3
Oct-2013
Latency - N. Gee 3Slide4
CommentsThe calculation shows remaining spare of ~ 10BC, but this is for Run 3 and probably should not match
Thilo’s Run 2 numberI am not sure that high-speed link latencies are as long as they need to be, including multiplexingThe multi-Gbit
signals we have running are typically 5BCs, some of the tables still have 4 BCs. Muon/Tile-D-layer interface is not yet included and needs to be addedOptional gFex is not included
Barrel Muon trigger is not well modelled, using measurements instead
(Riccardo is trying to get a breakdown)
Oct-2013Latency - N. Gee
4Slide5
MeasurementsI would like to get the calculation right
Opportunity while ATLAS is in pieces/being rebuiltAlready started discussions of measurements inside current L1Calo before and after the system is changedE.g. measure longest cable by TDR while Tilecal is disconnectedWill also ask Muon colleagues to help re-measure installed & new muon
componentsI am still concerned that I don’t know the whole story, I fear we will be close to the limit, real latencies are always longer than estimatesThere will not be another big optimisation like the one I have just doneWatch this space…
Oct-2013
Latency - N. Gee
5