Power and Energy Courtesy of Dr. Daehyun Lim@WSU, Dr. Harris@HMC, Dr. Shmuel Wimer@BIU and Dr. Choi@PSU http://csce.uark.edu +1 (479) 575-6043 yrpeng@uark.edu
The Chip is HOT Power consumption increases with the transistor count 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 2
Power and Energy Power is drawn from a voltage source attached to the V DD pin(s) of a chip. Instantaneous Power: P( t) I( t) V ( t) Energy: E T P() t dt 0 Average Power: P avg T E 1 T T 0 P() t dt A typical power breakdown 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 3
Power in Circuit Elements Voltage/Current Source: Power consumption: P t I t V Resistor: Heat dissipation: VDD DD DD t 2 VR 2 PR t I Rt R R Capacitor: Energy Stored: dv EC I tv tdt C V t dt dt 0 0 V C 0 C V t dv CV 1 2 2 C 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 4
Charging a Capacitor When the gate output rises Energy stored in capacitor is E C V 1 2 C 2 L DD But energy drawn from the supply is dv E I t V dt C V dt VDD DD L DD dt 0 0 VDD 2 L DD L DD 0 C V dv C V Half the energy from VDD is dissipated in the pmos transistor as heat, other half stored in capacitor When the gate output falls Energy in capacitor is dumped to GND Dissipated as heat in the nmos transistor 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 5
Switching Waveforms Example: V DD = 1.0 V, C L = 150 ff, f = 1 GHz 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 6
Switching Power Switch power calculation: T 1 Pswitching idd () t VDDdt T 0 V T DD V T CV DD T 0 i 2 DD DD sw sw () t dt Tf CV f DD VDD i DD (t) f sw C 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 7
Activity Factor Suppose the system clock frequency = f Let f sw = af, where a = activity factor If the signal is a clock, a = 1 If the signal switches once per cycle, a = ½ Dynamic power: P acv f switching 2 DD For dynamic power reduction, try to minimize: Activity factor Capacitance Supply voltage Frequency 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 8
Short Circuit Current When transistors switch, both nmos and pmos networks may be momentarily ON simultaneously Leads to a blip of short circuit current. < 10% of dynamic power if rise/fall times are comparable for input and output We will generally ignore this component 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 9
Power Dissipation Sources Total power: P total = P dynamic + P static Dynamic power: P dynamic = P switching + P shortcircuit Switching load capacitances Short-circuit current Static power: P static = (I sub + I gate + I junct + I contention )V DD Subthreshold leakage Gate leakage Junction leakage Contention current 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 10
Dynamic Power Example 1 billion transistor chip 50M logic transistors Average width: 12 l, Activity factor = 0.1 950M memory transistors Average width: 4 l, Activity factor = 0.02 1.0 V 65 nm process C = 1 ff/mm (gate) + 0.8 ff/mm (diffusion) Estimate dynamic power consumption @ 1 GHz. Neglect wire capacitance and short-circuit current. logic mem 6 l m l m C 5010 12 0.025 m / 1.8 ff / m 27 nf 6 l m l m C 95010 4 0.025 m / 1.8 ff / m 171 nf 2 Pdynamic 0.1C logic 0.02C mem 1.0 1.0 GHz 6.1 W 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 11
Activity Factor Estimation Let P i = Prob(node i = 1) P i = 1-P i a i = P i * P i Completely random data has P = 0.5 and a = 0.25 Data is often not completely random e.g. upper bits of 64-bit words representing bank account balances are usually 0 Data propagating through ANDs and ORs has lower activity factor Depends on design, but typically a 0.1 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 12
Switching Probability Activity of logic functions 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 13
Example A 4-input AND is built out of two levels of gates Estimate the activity factor at each node if the inputs have P = 0.5 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 14
Clock Gating The best way to reduce the activity is to turn off the clock to registers in unused blocks Saves clock activity (a = 1) Eliminates all switching activity in the block Requires determining if block will be used 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 15
Capacitance Gate capacitance Fewer stages of logic Small gate sizes Wire capacitance Good floorplanning to keep communicating blocks close to each other Drive long wires with inverters or buffers rather than complex gates 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 16
Voltage / Frequency Run each block at the lowest possible voltage and frequency that meets performance requirements Voltage Domains Provide separate supplies to different blocks Level converters required when crossing from low to high V DD domains Level shifter Dynamic Voltage Scaling Adjust V DD and f according to workload 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 17
Static Power Static power is consumed even when chip is quiescent. Leakage draws power from nominally OFF devices Ratioed circuits burn power in fight between ON transistors Example: Revisit power estimation for 1 billion transistor chip Estimate static power consumption Subthreshold leakage Normal V t : 100 na/mm High V t : 10 na/mm High Vt used in all memories and in 95% of logic gates Gate leakage 5 na/mm Junction leakage negligible 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 18
Solution W W normal-v high-v t t 6 6 50 10 12l 0.025mm / l 0.05 0.75 10 mm 6 6 6 50 10 12l 0.95 950 10 4l 0.025mm / l 109.25 10 mm Isub Wnormal-V 100 na/ mm+ W t high-v 10 na/ mm / 2 584 ma t I gate Wnormal-V W t high-v 5 na/ mm / 2 275 ma t P 584 ma static 275 ma1.0 V 859 mw 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 19
Subthreshold Leakage For Vds > 50 mv I sub I off 10 V V V k V gs ds DD sb S Ioff = leakage at Vgs = 0, Vds = VDD Typical values in 65 nm Ioff = 100 na/mm @ Vt = 0.3 V Ioff = 10 na/mm @ Vt = 0.4 V Ioff = 1 na/mm @ Vt = 0.5 V h = 0.1 kg = 0.1 S = 100 mv/decade 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 20
Stack Effect Series OFF transistors have less leakage Vx > 0, so N2 has negative Vgs V V x DD V V V V k V S I I 10 I 10 V sub off off x VDD 1 2 N2 N1 k Leakage through 2-stack reduces ~10x Leakage through 3-stack reduces further x DD x DD x 1 k VDD 1 2 k V S S sub off 10 off 10 I I I S DD 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 21
Leakage Control Leakage and delay trade off Aim for low leakage in sleep and low delay in active mode To reduce leakage: Increase V t : multiple V t Use low V t only in critical circuits (VTL/VTG/VTH in FreePDK45) Increase V s : stack effect Input vector control in sleep Decrease V b Reverse body bias in sleep (increase V t ) Or forward body bias in active mode (reduce V t ) 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 22
Gate Leakage Extremely strong function of t ox and V gs Negligible for older processes Approaches subthreshold leakage at 65 nm and below in some processes An order of magnitude less for pmos than nmos Control leakage in the process using t ox > 10.5 Å High-k gate dielectrics help Some processes provide multiple t ox e.g. thicker oxide for 3.3 V I/O transistors Control leakage in circuits by limiting V DD 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 23
NAND3 Leakage Example 100 nm process I gn = 6.3 na I gp = 0 I offn = 5.63 na I offp = 9.3 na 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 24
Junction Leakage From reverse-biased p-n junctions Between diffusion and substrate or well Ordinary diode leakage is negligible Band-to-band tunneling (BTBT) can be significant Especially in high-v t transistors where other leakage is small Worst at V db = V DD Gate-induced drain leakage (GIDL) exacerbates Worst for V gd = -V DD (or more negative) 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 25
Power Gating Turn OFF power to blocks when they are idle to save leakage Use virtual VDD (VDDV) Gate outputs to prevent invalid logic levels to next block Voltage drop across sleep transistor degrades performance during normal operation Size the transistor wide enough to minimize impact Switching wide sleep transistor costs dynamic power Only justified when circuit sleeps long enough 10/9/2017 CSCE/ELEG 4914: Advnaced Digital Design 26