Novita’ FTK da luglio P. Giannetti per il gruppo FTK • Review Amchip - lieve cambiamento alla schedule • Il problema del cooling ed I tests a punto 1 • Test miniasic e ordini di AMchip05 • Le schede • FTK nel TDR del TDAQ • Richieste e responsabilita’ Legenda Asic submission Stand-alone test Integrated test Global Integration test Production - Installation Review Months Tasks Dual Output HOLA FTK Input Mezzanine Data Formatter Miniasic AMchip05 AMchip06 AMBSLP-Mini-LAMBSLP AMBSLP-LAMBSLP AUX CARD Second Stage Board (SSB) FTK Level-2 Interface Crate (FLIC) cooling TDR schedule Review AMchip Review Boards 7 8 9 10 11 12 2013 1 2 3 4 Review cooling @point 1 5 w DF/IBL/RODs test w AUX test ta peout 6 7 8 9 10 11 12 2014 Global Int. Global Int. Global Int. test test test 1 2 3 4 2015 5 test tapeout test MOU schedule test aux w DF w FLIC/AUX w SSB - ROS w AM05AUX w AMBSLP Global Int. Global Int. Global Int. Global Int. wAM06 8-16 PUs 8-16 PUs 6 2 Gb/s data transfer • MiniAsic - ongoing tests at Milan – Silicon Creation Serializer/Deserializer OK • Design of AMchip05 (LPNHE-MI-LNF-PI-) advanced – Submission during october • Design of AMchip06 expected for spring 2014 – as early as possible to maintain the commissioning milestone BGA package common to AMchip05 & 06 defined – Amchip05 order started PACKAGE ASE DESIGN VSSS VSSS VSSS VSSS VSSS VSSS PATTIN1H OLD PATTIN0H OLD VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS PATTIN1_ P PATTIN1_ N VSSS VSSS VSSS VSSS VDDIO VDDIO VDDIO VDDH VSSS VSSS VDDIO VDDIO VDDIO VDDA VSS VDDIO VSS VSS VDDIO VSS VDDIO VDDIO VSS VDDIO VDDIO VSS VSS VSS VSS VSS VSS VDDIO VDDIO VDDIO VSS VSS VDDIO VDDIO VDDFC VDDFC VSS VSS VDDIO VDDIO VDDIO VSS VSS VDDIO VDDIO VDDIO VSS VSS VDDIO VDDIO VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS TMS TDI INIT TRST VSSS H1_P VSSS H3_P VSSS H5_P VSSS H7_P VSSS VSSS PATTOUT HOLD VSSS VSSS VSSS VSSS VSSS H1_N VSSS H3_N VSSS H5_N VSSS H7_N VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VDDIO VDDIO VSSS VSSS VDDIO VDDIO VSSS VSSS VDDIO VDDIO VDDIO VSS VDDIO VDDIO VSS VSS VDDIO VDDIO VSS VSS VSSS H0_P VSSS H2_P VSSS H4_P VSSS H6_P VSSS VSSS H0_N VSSS H2_N VSSS H4_N VSSS H6_N VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VDDH VDDH VSSS VSSS VSSS VSSS VSSS VSSS VSSS VDDH VDDA VDDA VDDH VDDH VDDH VDDH VDDH VDDH VDDH VDDA VSS VDDA VDDA VDDA VDDA VDDA VDDA VDDA VSS VSS VSSS VSSS VSSS VSSS VSSS VSSS VSSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VDDFC VDDFC VSS VSS VSS VSS VSS VSS VSS VSS VSS VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDSERD VDDSERD VDDSERD VSS ES ES ES VDDSERD VDDSERD VDDIO VSS ES ES VDDSERD VDDSERD VDDIO VDDIO ES ES VDDSERD VDDSERD VDDIO VDDIO ES ES VDDIO VDDIO VDDFC VDDFC VDDIO VDDFC VDDFC VDDFC VDDFC VDDFC VDDFC VDDFC VSS PATTIN0_ P PATTIN0_ N PATTOUT _P PATTOUT _N VSSS VSSS VSSS VSSS VSSS VSSS VDDSERD VDDH VDDH ES VDDSERD VDDA VDDA ES VDDSERD VDDSERD VSS ES ES VDDSERD VDDSERD VSS ES ES VDDSERD VDDSERD VDDIO ES ES VDDSERD VDDFC VDDIO ES VDDIO VDDIO VSS VSS VSS VSS VSS VSS VSS VDDFC VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS VSSA_LVD VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE S VDDFC VDDFC VDDFC VDDIO VDDFC VDDFC VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VSS VSS VSS VSS VSS VSS VDDFC VDDFC VDDIO VDDIO VSS VSS VDDIO VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDFC VDDFC VDDIO VDDIO VSS VSS VDDIO VDDCORE VDDCORE VDDCORE VDDIO VDDCORE VDDCORE VDDCORE VDDFC VDDFC VDDIO VDDIO VSS VSS VDDIO VDDIO VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDIO VSS VSS VDDIO VDDIO VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDIO VSS VSS VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VDDIO VSS VSS VDDIO VSS VSS VSS VSS VSS VSS VSS VSS VSS VSS TCK VSS VSS VSS VSS VSS VSS DTEST TDO VSS VSS VSSA_LVD S VSSA_LVD VDDCORE VDDCORE VDDCORE VDDCORE VDDIO VDDIO VDDIO S VSSA_LVD VSSA_LVD VDDCORE VDDIO VDDIO VDDIO VDDIO VDDIO S S VSSA_LVD VSSA_LVD VSSA_LVD VSS VSS VSS VSS VSS S S S VDDA_BG VSSA_BG VDDA_LV VSSA_LVD VSS VSS CLK_N CLK_P REF REF DS S VDDCORE VDDCORE VDDCORE VDDCORE VDDCORE VDDIO VDDIO AMCHIP status Responsabile A. Lanza (PAVIA) CAEN OPTION Rack Y.05-09.A2 SETUP for COOLING TESTs at Point 1 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 TURBINE HEAT EXCHANGER 9U VME bin (AM PU) Fan tray HEAT EXCHANGER External PS unit Fan tray HEAT EXCHANGER 9U VME bin (AM PU) Fan tray Closing panel Air deflector Network switch Rack Y.05-09.A2 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 TURBINE HEAT EXCHANGER External PS unit HEAT EXCHANGER 9U VME bin (AM PU) Fan tray Closing panel HEAT EXCHANGER External PS unit HEAT EXCHANGER 9U VME bin (AM PU) Fan tray Closing panel Air deflector Network switch BOARDS • FTK_IM prototype compatible with DF via FMC connector. • Problem with power generator solved (changed component) • Output to DF tested up to 400 MHz (design) for some lines. • New requirement from DF: more lines at 400MHz required → new version needed FRASCATI / WASEDA La nuova scheda AMBSLP just arrived – under test XILINX ARTIX 7 in place of SPARTAN 6 Smaller LAMBS → normal shape High Density connectors NO PARALLEL BUSES ONLY SERIAL LINK CONNECTIONS miniLAMBSLP and LAMBSLP miniLAMBSLP BGA 23x23 LAMBSLP READY to order QFN64 HW is OK but…. TDAQ - TDR….. • I samples dalla produzione ufficiale di ATLAS per rifare I plots di FTK con IBL incluso nella simulazione non sono ancora disponibili. • Integrazione con il sistema di produzione ha richiesto lavoro addizionale ed imprevisto, sample di calibrazione del sistema arrivati con 1 mese di ritardo • Esiste il capitolo 8 pronto su FTK, ma I plot sono vuoti → non possiamo essere sicuri di riuscire ad ordinare AMchip06 nel 2013 se approvazione FTK ritarda - proponiamo di spostare i fondi (200 keuro) al 2014 Richieste SBLOCCO SJ 2013: solo 5 k€ su Pisa (non core) per completare miniLAMB e LAMBSLP PER IL 2014: FRASCATI: Produzione mezzanine FTK_IM -> 135 keuro; CPU ATCA (non core) 7 keuro. PISA: schede e chips per 20 AMBSLPs -> 30 Pisa + 10 keuro SJ di contingenza. MILANO: 1. AMchip06 MLM masks 200 keuro (I 195 keuro recuperati dal 2013 + 5) 2. tests AMchips alla Microtest 50 keuro + 50 SJ alla definizione finale del costo. PAVIA: consumo per test di raffreddamento e sviluppo del controllo del power supply: 5 keuro TOT = [135 + 30 + 250] core + [7+5] (not core) = 415 k€ (core) + 12 k€ (non core) + 60 k€ SJ RESPONSABILITA’ ITALIANE IN FTK FTK team organization Deputy Project Manager - P. Giannetti (Pisa) Task LeadersHardware FTK_IM - M. Beretta (Frascati) AMBoard - M. Piendibene (Pisa) LAMB – P. Giannetti (Pisa) AMBoard and LAMB firmware - D. Magalotti (Perugia) AM chip - A. Stabile (Milan), System Integration Tests & board integration in the Vertical Slice - M. Piendibene (PI) DAQ integration: Vertical Slice/Demonstrator - A. Annovi (Frascati) Rack integration including power supplies, cooling, & safety - A. Lanza (Pavia) Interface to level-2 - A. Negri (Pavia), A. Annovi (Frascati) FTK simulation - G. Volpi (LNF) THIS IS THE PAST: project leader will be voted and will nominate all the others Conclusions • Hardware development is ok • The collaboration is strong with new important entries • Simulation and performance studies with IBL accumulated a significant delay MILESTONES "Test AMchip05 completo" milestone di fine giugno "Completa integrazione AMBSLP in FTK con AMchip05" milestone di dicembre. AMchip05 design • • • LVDS @ 2GHz: 11 SERDES (2 pattern in, 1 pattern out, 8 hit buses) LVDS @ 100 MHz: CLK single-ended control signals: JTAG Init, Dtest, Holds 14 Changes in LOGIC (LPNHE-Milan): o SERDES I/O @ 16 bits (2 DC) o o o o (AMchip04 was 15 bits 3 DC. Internally it's always 18 bits with configurable DC) Two pattern inputs one pattern output (merge of pattern streams) 1-layer match threshold (other thresholds: never, 8, 7, 6, always) double width mode (4 bus - 32 bit) optional continuous readout mode (AMchip04 was event based only) New features wrt AMchip04 Change for implementation of design o Majority inside pattern becomes full custom (MILAN) o New Low Power full custom cell for pattern (LNF) Milan 15 Not USA responsabilities cost sharing – 2014 -2017 – TDR status FTK Am & FTK_IM AMBSLP-LAMBSLP 80 '14; 80 '15; 390 '16-'17 DF mezzanine - ITA 2014 DF mezzanine - Waseda 2014 Am05 12 mm^2 MPW 2013 100 PBGA1 for AM05 2013 AMchip06 MLM masks 2013 eng lot 9 wafers for first AMchip06 production package PBGA for AMchip06 +40 AMchip06 production test ~50wafers: 2 lots of AMchip06 2015 - 2016-17 PBGA production 9000 pieces for AMchip06 SUPER TOT TOT k€ 550 135 135 45,00 18,00 410,00 55,00 20,00 120,00 300,00 35,00 1823,00 UNI-GE ITA 30 500 135 Waseda Heidelberg USA AM Melburne AUTH Parigi 10 10 STOT 135 45 0 20 200 190 xx 55 60 120 200 90 1175 100 35 170 100 290 ITALIA core 2013: = 220 keuro AMchip06 ITALIA core 2014: 135 (FTK_IM)+ 30 (AMBSLP) + 120 (tests parzialmente SJ) = 285 keuro ITALIA core 2015: 80 (AMBSLP)+200 wafers = 280 keuro ITALIS core 2016-2017: = 390 keuro AMBSLP + qualche spesa piccola non-core 10 10 1845 16