ESE370: Circuit-Level Modeling, Design, and Optimization for Digital Systems Day 12: October 3, 2011 Variation 1 Penn ESE370 Fall2011 -- DeHon Previously • Understand how to model transistor behavior • Given that we know its parameters – Vdd, Vth, tOX, COX, W, L, NA … CGC CGCS CGCB 2 Penn ESE370 Fall2011 -- DeHon But… • We don’t know its parameters (perfectly) 1.Fabrication parameters have error range 2.Identically drawn devices differ 3.Parameters change with environment (e.g. Temperature) 4.Parameters change with time (aging) Why I am more concerned with robustness than precision. 3 Penn ESE370 Fall2011 -- DeHon Today • Sources of Variation – Fabrication – Operation – Aging • Coping with Variation – Margin – Corners – Binning 4 Penn ESE370 Fall2011 -- DeHon Fabrication 5 Penn ESE370 Fall2011 -- DeHon Process Shift • • • • Oxide thickness Doping level Layer alignment Growth and Etch rates and times – Depend on chemical concentrations • How precisely can we control those? • Vary machine-to-machine, day-to-day • Impact all transistors on wafer 6 Penn ESE370 Fall2011 -- DeHon Systematic Spatial • Parameters change consistently across wafer or chip based on location • Chemical-Mechanical Polishing (CMP) – Dishing • Lens distortion 7 Penn ESE370 Fall2011 -- DeHon FPGA Systematic Variation • 65nm • Virtex 5 Penn ESE370 Fall2011 -- DeHon [Tuan et al. / ISQED 2010] 8 Oxide Thickness [Asenov et al. TRED 2002] 9 Penn ESE370 Fall2011 -- DeHon Line Edge Roughness • 1.2mm and 2.4mm lines From: http://www.microtechweb.com/2d/lw_pict.htm 10 Penn ESE370 Fall2011 -- DeHon Optical Sources • What is the shortest wavelength of visible light? • How compare to 45nm feature size? 11 Penn ESE370 Fall2011 -- DeHon Phase Shift Masking Today’s chips use λ=193nm Source http://www.synopsys.com/Tools/Manufacturing/MaskSynthesis/PSMCreate/Pages/default.aspx 12 Penn ESE370 Fall2011 -- DeHon Line Edges (PSM) Source: http://www.solid-state.com/display_article/122066/5/none/none/Feat/Developments-in-materials-for-157nm-photoresists 13 Penn ESE370 Fall2011 -- DeHon Intel 65nm SRAM (PSM) Source: http://www.intel.com/technology/itj/2008/v12i2/5-design/figures/Figure_5_lg.gif 14 Penn ESE370 Fall2011 -- DeHon Statistical Dopant Placement 15 Penn ESE370 Fall2011 -- DeHon [Bernstein et al, IBM JRD 2006] Random Trans-to-Trans • • • • Random dopant fluctuation Local oxide variation Line edge roughness Etch and growth rates – Stochastic process • Transistors differ from each other in random ways Penn ESE370 Fall2011 -- DeHon 16 Source: Noel Menezes, Intel ISPD2007 17 Penn ESE370 Fall2011 -- DeHon Impact • Changes parameters – W, L, tOX, Vth • Change transistor behavior – W? – L? – tOX? IDS mn COX W 2 VGS VT L 2 18 Penn ESE370 Fall2011 -- DeHon Example: Vth • Many physical effects impact Vth – Doping, dimensions, roughness • Behavior highly dependent on Vth IDS IDS 2 W IS e L Penn ESE370 Fall2011 -- DeHon mn COX W VGS VT L 2 VGS VT nkT / q VDS kT / q 1 e 1 VDS 19 Vth Variability @ 65nm 20 Penn ESE370 Fall2011 -- DeHon [Bernstein et al, IBM JRD 2006] Impact of Vth Variation? • Higher VTH? – Not drive as strongly – Id,sat (Vgs-VTH)2 – Performance? IDS mn COX W 2 VGS VT L 2 21 Penn ESE534 Spring2010 -- DeHon Impact Performance • Vth Ids Delay (Ron * Cload) 22 Penn ESE370 Fall2011 -- DeHon Impact of Vth Variation Penn ESE370 Fall2011 -- DeHon Think NMOS Vgs = Vdd 23 FPGA Logic Variation • Xilinx Virtex 5 • 65nm • Altera Cyclone-II • 90nm [Tuan et al. / ISQED 2010] Penn ESE370 Fall2011 -- DeHon [Wong, FPT2007] 24 Impact of Vth Variation? • Lower VTH? – Not turn off as well leaks more IDS W IS e L Penn ESE534 Spring2010 -- DeHon VGS VT nkT / q VDS kT / q 1 e 1 VDS 25 2004 Borkar (Intel) Micro 37 (2004) Penn ESE370 Fall2011 -- DeHon 26 Operation Temperature Voltage 27 Penn ESE370 Fall2011 -- DeHon Temperature Changes • Different ambient environments – January in Maine – July in Philly – Air conditioned machine room • Self heat from activity of chip • Quality of heat sink (attachment thereof) IDS W IS e L Penn ESE370 Fall2011 -- DeHon VGS VT nkT / q VDS kT / q 1 e 1 VDS 28 Self Heating Borkar (Intel) Micro 37 (2004) 29 Penn ESE370 Fall2011 -- DeHon Temperature (on current?) • High temperature – More free thermal energy • Easier to conduct • Lowers Vth – Increase rate of collision • Lower saturation velocity • Lower saturation voltage • Lower peak Ids slows down • Another reason don’t want chips to run hot 30 Penn ESE370 Fall2011 -- DeHon Temperature (leakage?) • High temperature Lowers Vth IDS W IS e L Penn ESE370 Fall2011 -- DeHon VGS VT nkT / q VDS kT / q 1 e 1 VDS 31 Voltage • Power supply isn’t perfect • Differs from design to design – Board to board? – How precise is regulator? • IR-drop in distribution • Bounce with current spikes 32 Penn ESE370 Fall2011 -- DeHon Aging Hot Carrier NBTI 33 Penn ESE370 Fall2011 -- DeHon Hot Carriers • Trap electrons in oxide – Also shifts Vth 34 Penn ESE370 Fall2011 -- DeHon NBTI • Negative Bias Temperature Instability – Interface traps, Holes • Long-term negative gate-source voltage – Affects PFET most • Increase Vth • Partially recoverable? • Temperature dependent Penn ESE370 Fall2011 -- DeHon [Stott, FPGA2010] 35 Measured Accelerated Aging (Cyclone III, 65nm FPGA) Penn ESE370 Fall2011 -- DeHon [Stott, FPGA2010] 36 Coping with Variation 37 Penn ESE370 Fall2011 -- DeHon Variation • See a range of parameters – L: Lmin – Lmax – Vth: Vth,min – Vth,max 38 Penn ESE370 Fall2011 -- DeHon Impact of Vth Variation • Higher VTH – Not drive as strongly – Id,sat (Vgs-VTH)2 • Lower VTH – Not turn off as well leaks more IDS W IS e L Penn ESE534 Spring2010 -- DeHon VGS VT nkT / q VDS kT / q 1 e 1 VDS 39 Variation • Margin for expected variation • Must assume Vth can be any value in range – Speed assume Vth slowest value Ion,min=Ion(Vth,max) Probability Distribution Id,sat (Vgs-Vth)2 VTH 40 Penn ESE370 Fall2011 -- DeHon Gaussian Distribution From: http://en.wikipedia.org/wiki/File:Standard_deviation_diagram.svg 41 Penn ESE534 Spring2010 -- DeHon Impact • Given – Vth,nom = 250mV – Sigma 25mV • What maximum Vth should expect to see for a circuit of – 100 transistors? – 1000 transistors? – 109 transistors? 42 Penn ESE370 Fall2011 -- DeHon Variation • See a range of parameters – L: Lmin – Lmax – Vth: Vth,min – Vth,max • Validate design at extremes – Work for both Vth,min and Vth,max ? – Design for worst-case scenario 43 Penn ESE370 Fall2011 -- DeHon Margining • Also margin for – Temperature – Voltage – Aging: end-of-life 44 Penn ESE370 Fall2011 -- DeHon Process Corners • Many effects independent • Many parameters • With N parameters, – Look only at extreme ends (low, high) – How many cases? • Try to identify the {worst,best} set of parameters – Slow corner of design space, fast corner • Use corners to bracket behavior Penn ESE370 Fall2011 -- DeHon 45 Range of Behavior Probability Distribution • Still get range of performances • Any way to exploit the fact some are faster? Delay 46 Penn ESE370 Fall2011 -- DeHon Probability Distribution Speed Binning Sell Premium Sell Sell nominal cheap Discard Delay 47 Penn ESE370 Fall2011 -- DeHon Admin • HW4 out – All but question 5 should be approachable now • Andre out Tuesday • Andre back for lecture on Wednesday – Talk about layout … HW4.Q5 • First midterm W after break 48 Penn ESE370 Fall2011 -- DeHon Idea • Parameters Approximate • Differ – Chip-to-chip, transistor-to-transistor, over time • Robust design accommodates – Tolerance and Margins – Doesn’t depend on precise behavior Penn ESE370 Fall2011 -- DeHon 49