The Matrix Card and its Applications

Slides:



Advertisements
Similar presentations
Digital RF Stabilization System Based on MicroTCA Technology - Libera LLRF Robert Černe May 2010, RT10, Lisboa
Advertisements

ESODAC Study for a new ESO Detector Array Controller.
Uli Schäfer 1 S-L1Calo upstream links architecture -- interfaces -- technology.
Uli Schäfer 1 FPGAs for high performance – high density applications Intro Requirements of future trigger systems Features of recent FPGA families 9U *
C. Foudas, Imperial CMS Upgrades Workshop FNAL – Trigger Introduction Status of the CMS Trigger Upgrade Activities: –uTCA and Optical Link upgrades.
The Track-Finding Processor for the Level-1 Trigger of the CMS Endcap Muon System D.Acosta, A.Madorsky, B.Scurlock, S.M.Wang University of Florida A.Atamanchuk,
CPT Week, April 2001Darin Acosta1 Status of the Next Generation CSC Track-Finder D.Acosta University of Florida.
Upgrade of the CSC Endcap Muon Port Card Mikhail Matveev Rice University 1 November 2011.
The L0 Calorimeter Trigger U. Marconi On behalf of the Bologna Group CSN1, Catania 16/9/02.
W. Smith, DOE/NSF Review, August, 2006 CMS Trigger - 1 SORT ASICs (w/heat sinks) EISO Bar Code Input DC-DC Converters Clock delay adjust Clock Input Oscillator.
FED RAL: Greg Iles5 March The 96 Channel FED Tester What needs to be tested ? Requirements for 96 channel tester ? Baseline design Functionality.
ATLAS Trigger / current L1Calo Uli Schäfer 1 Jet/Energy module calo µ CTP L1.
01/04/09A. Salamon – TDAQ WG - CERN1 LKr calorimeter L0 trigger V. Bonaiuto, L. Cesaroni, A. Fucci, A. Salamon, G. Salina, F. Sargeni.
US CMS DOE/NSF Review, May Cal. Trig. 4 Gbaud Copper Link Cards & Serial Link Test Card - U. Wisconsin Compact Mezzanine Cards for each Receiver.
W. Smith, U. Wisconsin, US CMS DOE/NSF Review, September 2004 Trigger Report - 1 CSC Trigger Muon Port Card & Sector Processor in production Mezzanine.
US CMS DOE/NSF Review, 20 November Trigger - W. Smith1 WBS Trigger Wesley H. Smith, U. Wisconsin CMS Trigger Project Manager DOE/NSF Status.
Trigger Hardware Development Modular Trigger Processing Architecture Matt Stettler, Magnus Hansen CERN Costas Foudas, Greg Iles, John Jones Imperial College.
Trigger Workshop: Greg Iles Feb Optical Global Trigger Interface Card Dual CMC card with Virtex 5 LX110T 16 bidirectional.
W. Smith, U. Wisconsin, US CMS DOE/NSF Review, May, 2004 Trigger Report - 1 CSC on-detector peripheral crate SBS VME Controller Muon Port Card: Output.
1 Timing of the calorimeter monitoring signals 1.Introduction 2.LED trigger signal timing * propagation delay of the broadcast calibration command * calibration.
S. Dasu, University of Wisconsin February Calorimeter Trigger for Super LHC Electrons, Photons,  -jets, Jets, Missing E T Current Algorithms.
August 24, 2011IDAP Kick-off meeting - TileCal ATLAS TileCal Upgrade LHC and ATLAS current status LHC designed for cm -2 s 7+7 TeV Limited to.
E. Hazen - DTC1 DAQ / Trigger Card for HCAL SLHC Readout E. Hazen - Boston University.
TBD The Matrix Card Status and Plans Dr. John Jones Princeton University
29/05/09A. Salamon – TDAQ WG - CERN1 LKr calorimeter L0 trigger V. Bonaiuto, L. Cesaroni, A. Fucci, A. Salamon, G. Salina, F. Sargeni.
The LHCb Calorimeter Triggers LAL Orsay and INFN Bologna.
E. Hazen1 MicroTCA for HCAL and CMS Review / Status E. Hazen - Boston University for the CMS Collaboration.
E. Hazen -- ACES ACES 2011 MicroTCA in CMS E. Hazen, Boston University for the CMS collaboration.
E. Hazen -- xTCA IG1 AMC13 Project Status E. Hazen - Boston University for the CMS Collaboration.
E. Hazen -- Upgrade Week1 HCAL uTCA Readout Plan E. Hazen - Boston University for the CMS HCAL Collaboration.
Krzysztof Czuba, ISE, Warsaw ATCA - LLRF project review, DESY, Dec , XFEL The European X-Ray Laser Project X-Ray Free-Electron Laser Master Oscillator.
P. Jansweijer Nikhef Amsterdam Electronics- Technology Amsterdam July 5-6, 2010KM3NeT: General WPF/L meeting 1 Measuring time offset over a bidirectional.
E. Hazen1 Fermilab CMS Upgrade Workshop November 19-20, 2008 A summary of off-detector calorimeter TriDAS electronics issues Eric Hazen, Boston.
E. Hazen -- Upgrade Week1 MicroTCA Common Platform For CMS Working Group E. Hazen - Boston University for the CMS Collaboration.
Results with the RPC system of OPERA and perspectives
AMC13 Project Status E. Hazen - Boston University
DAQ and TTC Integration For MicroTCA in CMS
AMC13 T1 Rev 2 Preliminary Design Review E. Hazen Boston University
D. Breton, S. Simion February 2012
ATLAS calorimeter and topological trigger upgrades for Phase 1
CEPC 数字强子量能器读出电子学预研进展
Future Hardware Development for discussion with JLU Giessen
2018/6/15 The Fast Tracker Real Time Processor and Its Impact on the Muon Isolation, Tau & b-Jet Online Selections at ATLAS Francesco Crescioli1 1University.
A Compact Trigger Architecture for SLHC
L1Calo upgrade discussion
Update on CSC Endcap Muon Port Card
TELL1 A common data acquisition board for LHCb
uTCA A Common Hardware Platform for CMS Trigger Upgrades
Trigger Upgrade Planning
CMS EMU TRIGGER ELECTRONICS
ATLAS L1Calo Phase2 Upgrade
MicroTCA Common Platform For CMS Working Group
TCC-48 project status report
Status of the oSLB project
CMS SLHC Calorimeter Trigger Upgrade,
ECAL OD Electronic Workshop 7-8/04/2005
Implementation of the Jet Algorithm ATLAS Level-1 Calorimeter Trigger
xTCA in CMS Magnus Hansen Thanks to many CMS collaborators
Calorimeter Trigger Synchronization in CMS,
Matrix Processor / Backplane Design Details
Regional Cal. Trigger Milestone: Major Production Complete
Regional Cal. Trigger Milestone: Production & Testing Complete
CSC Trigger Muon Port Card & Sector Processor in production
Jason Gilmore Vadim Khotilovich Alexei Safonov Indara Suarez
Measuring propagation delay over a coded serial communication channel using FPGAs P.P.M. Jansweijer, H.Z. Peek October 15, 2009 VLVnT-09 Athens.
The LHCb L0 Calorimeter Trigger
TELL1 A common data acquisition board for LHCb
(Not just) Backplane transmission options
Fixed Latency Serial Links with FPGA-embedded SerDes for SuperB
LIP and the CMS Trigger Upgrade On behalf of the LIP CMS Group
Presentation transcript:

The Matrix Card and its Applications TBD FERMI ELETTRA The Matrix Card and its Applications 23.09.2009 Dr. John Jones Princeton University neutrinodeathray@gmail.com CMS CERN

Progression in Physics Hardware Over ~20 Years Speed Connectivity Density Size 1995 VME 2003 VME / ATCA 2008+ ATCA / μTCA 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

The Matrix Processor - Schematic SNAP12 2Gb DDR2 Xilinx Virtex 5 FPGA (LX110T) Mindspeed 72x72 Cross-point Switch POP4 OPTICAL I/O (16/16) 3U μTCA I/O (20/20) 2Gb DDR2 SNAP12 NXP 2366 μC 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

The Matrix Processor – Top Photo FPGA MTP Optics Ethernet Mindspeed Switch 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

The Matrix Processor – Bottom Photo NXP Microcontroller DDR2 SDRAM TCA Connector 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Switched Topology Processing The switch technology makes it topologically agnostic: This is a critical difference compared to previous systems Allows the system to be used to solve calculations in various geometries 1-to-N data duplication is easy to implement w/ latency ~100ps Also allows real-time redundancy, dynamic switching, etc… Linear (e.g. batch) 3D (e.g. lattice sim.) 2D / projective 2D (e.g. CMS) 4D… 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Example 1 – CMS Trigger Upgrade 2-phase upgrade of trigger system: Phase 1 (2011  2016): Replacement of older components, HCAL FE & associated trigger hardware Calorimeter trigger upgrade Phase 2 (Some time after...): Installation of upgraded tracker including TP generation Integration of tracking information into enhanced trigger system 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Dr. John Jones (neutrinodeathray@gmail.com) Collision t=0 0.5 ECAL FE (4.5 / 4.5) HCAL FE (? / ?) 19* 0.8 ? 1.6 ECAL TCC (17 / 17*) HCAL HTR (? / ?) 5* 1.2 4* Current CMS Trigger Architecture RCT (20 / 18*) 0.08 1.5 External link latency (BX) Link speed (Gb/s) Internal processing latency (with internal connections) Internal processing latency (without internal connections) GCT (25 / 9) 3.0 5.5 GT (11 / 6) 51.5 (37) 79.5 56.5 0.04 1 TTC (2* / 2*) 19* Detector readout t=131 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Current CMS Trigger Architecture Processing subdivided into eta-phi regions / link (e.g. calorimeter trigger) 2 scaling problems with this approach: Difficult to add new input sources (i.e. improved HCAL, tracking) Data reduction layer doesn’t scale efficiently & balance boundary data sharing CAL TPG φ 400 RCT 20 GCT η 1 Significant data reduction GT 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Current-Revised CMS Trigger Architecture Revisit calorimeter TPG principle: Current Revised (time-mulitplexed serialisation) CAL TPG CAL TPG CAL TPG CAL TPG CAL TPG CAL TPG CAL TPG RCT CAL TPG CAL TPG ROG CAL TPG ηφ=1, t=1 ηφ=3, t=3 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Data Serialisation in TPG TPG multiplexes data into BX-serialised streams: φ t2 φ t1 t0 η Overall latency DECREASES! φ t0 t1 t2 Initial cost: lost time due to multiplexing Later gain: Compact, redundant, time-multiplexed system up to GT 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Current CMS Trigger Architecture Processing subdivided into eta-phi regions / link (e.g. calorimeter trigger) CAL TPG MUON TPG 400 Region / card increases Eliminates GCT / RCT boundary Space for additional future data Inter-card data sharing decreases ROG 20 GT 1 More compact Faster Lower latency Topological OR 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Doing the Numbers (Based on Current CT) Post-TPG link speed ~3.75Gb/s ~8b * 9.375 / BX / fibre 16 x serialisation in TPG => ~75 towers (ECAL+HCAL) / BX / fibre Eliminate phi-boundary (one fibre absorbs entire eta segment!) Calorimeter dimensions 88 (eta) x 72 (phi) trigger towers e.g. 1 matrix card = 16 (eta) x 72 (phi) 16 input channels => all inputs for jet trigger + overlap in current CMS 10 matrix cards for full-phi-granularity, coarse (4 tower) eta processing (x16 copies) 16 matrix cards for full-tower-granularity processing (x16 copies) 2 fibers => output for results (electrons, photons & jets) 32 input fibres into GT card 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Processing Topology – New and Old η New Scheme 3x3 jet tower finder (full phi resolution) 4x4 calorimeter towers / jet tower 3.75Gb/s links 22x18 (88x72) Data sharing – input fiber ratio: 160/88 = 1.82 Fibers Processing Real input fiber count: 16x88 = ~1408 φ η Old scheme – NN sharing 6.5Gb/s links Data sharing – input fiber ratio: ~21888/680 = 32.19 Real input fiber count: 16x72x88/144 = ~680 φ Factor of two from link speed – need 6.5Gb/s to use old scheme 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

The Modular Trigger Crate – 3.5Gb/s, Partial Granularity Can have a fully-redundant crate (spare fibres from TPG) Redundant power & communications Improvements in link speed = reduction in crate size or latency Complete system test can be achieved with a small setup (e.g. debug) PWR1 CMS AUX/DAQ MINI-T 12 8 8 8 MCH1 8 8 8 8 8 12 PWR2 MATRIX MATRIX MATRIX MATRIX MCH2 MATRIX MATRIX MATRIX MATRIX MATRIX MATRIX DATA 20 CLK 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Example 2 – FERMI, Trieste 4th generation Free Electron Laser (FEL) http://www.elettra.trieste.it/FERMI Linear accelerator, VUV-XRAY (10-100nm) Extremely challenging (3GHz) RF control system Tolerance: 0.1% amplitude, 0.1 degree phase Precision (~20fs accuracy over 24 hours / 200m distance) RF timing system Control / diagnostic system for RF cavities will use matrix card and LLRF board Control system accuracy: 50ps clock resolution, synchronised at 16 stations This will be achieved without a dedicated timing interface 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Timing System Principle A standard optical fiber has very similar path lengths (~ps) in each direction. Any change in path length in fiber of a TX/RX pair is matched by the other. If you have a timing reference at each end of a serial link with guaranteed constant phase relationship between them, you can measure the loop time and use it to measure the propagation delay from the master board (matrix) to the slave (LLRF), and therefore compensate for the delay. Such guaranteed phase can be achieved by either: 1) A shared reference clock at both ends of a link. 2) An extremely accurate OCXO that can be used to track the recovered serial clock at the receiving end. Given the available components in a Xilinx FPGA, the loop time can be measured consistently to an accuracy of ~50ps (Xilinx DCM limited). (N.B. With a few tricks, you can possibly do better) 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Round-Trip Phase Compensation, Version I RFCLK MATRIX LLRF δT +δFTPCBi δR +δNRPCBi TXCLK CLK BRIDGE GTP RX GTP TX TXCLK DCM δc δFL TXCLK δCB1 CLK BRIDGE CTRL GTP TX GTP RX δFCB δR +δFRPCBi δT +δNTPCBi LLRFCLK δCB2 CLK BRIDGE LLRFDATA DCM PIPELINE DELAY δFDCM NOTE: Control logic not shown 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Round-Trip Phase Compensation, Version II RFCLK MATRIX LLRF DPLL OCXO δT +δFTPCBi δR +δNRPCBi TXCLK CMP CLK BRIDGE GTP RX GTP TX TXCLK RXRECCLK δc δFL TXCLK δCB1 CLK BRIDGE CTRL GTP TX GTP RX δFCB δR +δFRPCBi δT +δNTPCBi LLRFCLK δCB2 CLK BRIDGE LLRFDATA DCM PIPELINE DELAY δFDCM NOTE: Control logic not shown 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Dr. John Jones (neutrinodeathray@gmail.com) Timing System Details Version I has been implemented, mostly finished (calibration in software a.t.m.) Caveat: 1 serial time UI variability seen on one channel in matrix card This needs further study, hard to reproduce and doesn’t occur on all channels Possible to correct for this effect using a loopback technique Xilinx datasheet implies this is an artifact of the way V5 MGTs work… …but they don’t tell you the details… Backup: Use LVDS @ 1Gb/s, which has completely deterministic behaviour 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)

Dr. John Jones (neutrinodeathray@gmail.com) Conclusions The Matrix Card is an extremely flexible device with many applications A large part of this flexibility comes from evolution in FPGA technology The addition of the cross-point switch provides significant extra flexibility 23.09.2009 Dr. John Jones (neutrinodeathray@gmail.com)