Basics of Energy-Efficient Design

Lin ZhongELEC424, Fall 2010

1

Outline

• General concepts• Energy-saving mechanisms• Integrated circuits (IC)

– Processors

• Wireless interfaces

2

What is energy efficiency?

• Power• Energy (Power * delay)• Battery lifetime• Energy * delay

3

Source of energy consumption

• IC– Computing/Switching– Radiation (wireless)

• Discrete components

• Display

4

Energy characterization

• Equipment setup

5

Power Supply

- +

Target

0.1 ohmSense Resistor

Voltage Sampling Device

6

Differential measurement

Extra energy/power consumption of an event obtained through differential measurements

Extra energy consumption for writing “x”

Write “x” with stylus/touchscreen

0

0.4

0.8

1.2

1.6

0 0.5 1 1.5

Time (s)

Pow

er (W

)

0

0.4

0.8

1.2

1.6

0 0.5 1 1.5

Time (s)

Pow

er (W

)

Rule No. 1• Focus on the big one!

– Amdahl’s Law• Reduction of the power of α % of the system by p% leads to

α∙p % reduction of the total power

7

α

Rule No. 2

• Minimize static energy consumption– IC consumes static power when it is merely

powered

8

Rule No. 3

• Minimize activity– When not doing things useful

• Turn it off• Stop the clock• Check the manual for power-saving modes• Be aware of state transition overhead

– Interrupt-driven instead of polling

9

Rule No. 4

• Don’t forget parasites – More integrated solution leads to lower energy

10

Processors

• Dynamic voltage scaling• Power-saving states

– Clock gating– Power-down different subsets of components

• As-fast-as-possible or As-slow-as-possible?

11

Power consumption of processing

• Dynamic power

12

Busy power vs. delay vs. energy

fVCaP dddyn ⋅⋅⋅= 2

)( Tdd

dd

VVVt−

∝

Analysis and Design of Digital ICs, Hodges et al

13

Core 2 Duo for example• Intel® Core™2 Duo processor

– T7800 at 2.6GHz– T7700 at 2.4GHz available on Thinkpad T61p– 0.75-1.35V, 35Watts

• Intel® Core™2 Duo Low Voltage– L7500 at 1.6GHz available on Thinkpad X61– 0.75-1.3V, 17Watts

• Intel® Core™2 Duo Ultra Low Voltage– U7500 at 1.06GHz available on Dell D430– 0.75-0.975V, 10Watts

14

Switching energy

e=1/2∙C ∙V2

Switching power

P= b∙C ∙V2= a∙C ∙V2 ∙f15

Given workload L and deadline T

• L measured by # of CPU cycles• Clock speed f ≥ L/T

• Time to finish: t = L/f

• Energy to finish: P ∙ t= a∙C ∙V2 ∙f ∙t= a∙C ∙V2 ∙L

16

Effect of lower clock speed (f)

Power consumption

P= a∙C ∙V2 ∙f

Energy consumption

E=P ∙ t= a∙C ∙V2 ∙f ∙t= a∙C ∙V2 ∙L

17

Effect of lower supply voltage (V)

Power consumption

P= a∙C ∙V2 ∙f=k∙V3=x∙f3

Energy consumption

E=P ∙ t= a∙C ∙V2 ∙f ∙t= a∙C ∙V2 ∙L

Maximum clock speed

f= b∙V

18

Given workload L and deadline Tsingle processor

• The processor can run at any frequency (voltage)– f= b∙V

• The processor can be complete off when work is done (zero power when idle)

• To minimize energy consumption, at which frequency should the processor run?– f ≥ L/T (in order to meet the deadline)– E=P ∙ t= a∙C ∙V2 ∙f ∙t= a∙C ∙V2 ∙L– f=????

19

time

f

T

f1=L/T

f2=L/(T/2)=2f1

20

time

P

T

P1=x∙f3

P2=23P1

21

Given workload L and deadline TM processors

• The workload can be divided without overhead: L = L1+L2+…+LM (L ≥ Li≥0)

• To minimize energy consumption, at which frequency should processor i run?– f i= Li/T and V = u ∙ Li

– Ei= a∙C ∙V2 ∙Li=w∙Li3

22

Given workload L and deadline TM processors

• The workload can be divided without overhead: L = L1+L2+…+LM (L ≥ Li≥0)

• To minimize the TOTAL energy consumption, how should the workload be allocated?– E= E1+E2+…+EM= w∙L1

3+w∙L23+…+w∙LM

3

– = w(L13+L2

3+…+LM3)

23

From high school

• [(a+b)/2]2≤ (a2+b2)/2

≥ ≥ ≥

Quadratic mean Arithmetic mean Geometric mean harmonic mean

24

From high school (Contd.)

• [(a+b)/2]3≤ (a3+b3)/2 ( for a, b ≥0)

– E= w(L13+L2

3+…+LM3) ??? (L1+L2+…+LM)3

25

From college: Convex (Concave)

By definition of “convex”

26

Jensen’s Inequality (finite form)

• ϕ (x) is convex– ϕ (t∙x1+(1-t)∙x2)≤ t∙ ϕ (x1)+(1-t) ∙ϕ (x2)

http://en.wikipedia.org/wiki/Jensen%27s_inequality#Proof_1_.28finite_form.29

27

• ai=1/n• ϕ (x) =x2 (Convex)

• ϕ (x) =x3(Convex for x≥0)– E= w(L1

3+L23+…+LM

3)=w∙M ∙ (L13+L2

3+…+LM3)/M

– ≥ w∙M ∙[(L1+L2+…+LM)/M] 3=w∙L3/M2

≥

28

Homework

• Use Jensen’s Inequality to prove

– for ai>0

≥

29

Check the assumptions

• Power consumption is zero when the processor is not active

30

Idle power (Static power)

)( Tdd

dd

VVVt−

∝ Tstatic eTP

α−

∝ 2 ddVddstatic eVP ⋅∝ γ

When IC is idle but not powered off, e.g. SRAM31

Multiple power/clock domains

TI OMAP 2 architecture, ISSCC 2005

Multimedia phone: NTT DoCoMo 3G FOMA 902ito be released with OMAP2420

32

time

f

T

f1=L/T

f2=L/(T/2)=2f1

33

time

P

T

P1=x∙f3

34

time

P

T

P1=x∙f3+Pstatic

35

time

P

T

P1=x∙f3+Pstatic

P2=23x∙f3+Pstatic

36

Given workload L and deadline Tsingle processor

• One processor can run at any frequency (voltage)– f= b∙V

• The processor can be complete off when work is done (zero power when idle) Given Pstatic– Given energy overhead of shutting down the

processor (Eoverhead)• To minimize energy consumption, at which

frequency should the processor run?

37

Check the assumptions (Contd.)

• The workload can be divided without overhead: L = L1+L2+…+LM (L ≥ Li≥0)

• Communication cost between processors!!!

38

Wireless interfaces

• Stay connected

• Establish connection

• Transfer data

• Transmit vs. receive

39

Energy per bit transfer

Oppermann et al., IEEE Comm. Mag. 200440

41

Power consumption (SMT5600)

Lighting: Keyboard, 73, 3% Lighting: Display I,

148, 5% Lighting: Display II, 61, 2%

LCD, 13, 0%

Speaker, 45, 2%

Bluetooth, 440, 16%

GPRS, 1600, 58%

Compute, 370, 13%

Cellular network, 17, 1%

Flight mode: Sleep, 3, 0%

42

Power consumption (T-Mobile)

1

10

100

1000

10000

IDLE-Flight m

ode

Com

puting

LCD

LCD

lighting

Keyboard lighting

Speaker

Discoverable

Paging

Connected

Transmission

Connected

Transmission

Connected

Transmission

Pow

er (m

W)

Bluetooth Wi-Fi Cellular

43

Power consumption (Contd.)

• Theoretical limits– Receiving energy per bit > N * 10-0.159

• N: Noise spectral power level• Wideband communication

Distance: d

Propagation constant: a (1.81-5.22)

PRXPTX∝ PRX*da

44

Power consumption (Contd.)

• What increases power consumption– National regulation (FCC)

• Available spectrum band (Higher band, higher power)• Limited bandwidth• Limited transmission power

– Noise and reliability– Higher capacity

• Multiple access (CDMA, TDMA etc.)– Security– Addressability (TCP/IP)– More……

Wasteful wireless communication

45

TimeMicro power management

SpaceDirectional communication

SpectrumEfficiency-driven cognitive radio

Space waste

• Omni transmission huge power by power amplifier (PA)

46

Time waste

• Network Bandwidth Under-Utilization– Modest data rate required by applications

• IE ~ 1Mbps, MSN video call ~ 3Mbps– Bandwidth limit of wired link

• 6Mbps DSL at home

47470 0.2 0.4 0.6 0.8 1

0

200

400

600

800

1000

1200

1400

Time (s)

Dat

a S

ize

(Byt

e)

0

20

40

60

80

100

Time Energy

Idle

inte

rval

s in

busy

tim

e (%

)

User1 User2 User3 User4

Spectrum waste

48

49

Wireless system architecture

Application

Transport

Network

Data link

Host computer

RF front ends

BasebandNetwork interface

Network protocol stack Hardware implementation

Physical

50

Power consumption (Contd.)

Baseband processor

Antenna interface

LNA

Low-noise amplifier

PA

Power amplifier

Intermediate Frequency (IF) signal processing

Local Oscillator (LO)

Physical Layer

IF/B

aseb

and

Conv

ersio

n

MAC Layer & above

>60% non-display power consumed in RF

RF technologies improve much slower than IC

51

Power consumption (Contd.)

67%

18%

8%

6%

1%

PA

FS

Mixer

Source: Li et al, 2004

Components Power (mW)

Power amplifier (PA) 246

Frequency synthesizer (VCO/FS)

67.5

Mixer 30.3

LNA 20

Baseband processing 5

52

Circuit power optimization

• Major power consumers

Baseband processor

Antenna interface

LNA

Low-noise amplifier

High duty cycle

PA

Power amplifier

High power consumption

Intermediate Frequency (IF) signal processing

Local Oscillator (LO)

Almost always on

Physical Layer

IF/B

aseb

and

Conv

ersio

n

MAC Layer & above

Huge dynamic range 105

53

Circuit power optimization (Contd.)

• Reduce supply voltage– Negatively impact amplifier linearity

• Higher integration– CMOS RF– SoC and SiP integration

• Power-saving modes

54

Circuit power optimization (Contd.)

• Power-saving modes– Complete power off

• (Circuit wake-up latency + network association latency) on the order of seconds

– Different power-saving modes• Less power saving but short wake-up latency

55

Power-saving modes

Baseband processor

Antenna interface

LNA

Low-noise amplifier

PA

Power amplifier

Intermediate Frequency (IF) signal processing

Local Oscillator (LO)

Physical Layer

IF/B

aseb

and

Conv

ersio

n

MAC Layer & above

Radio Deep Sleep Wake-up latency on the order of micro seconds

56

Power-saving modes (Contd.)

Baseband processor

Antenna interface

LNA

Low-noise amplifier

PA

Power amplifier

Intermediate Frequency (IF) signal processing

Local Oscillator (LO)

Physical Layer

IF/B

aseb

and

Conv

ersio

n

MAC Layer & above

Sleep Mode Wake-up latency on the order of milliseconds

Low-rate clock with saved network association information

57

Network power optimization

• Use power-saving modes– Example: 802.11 wireless LAN (WiFi)

• Infrastructure mode: Access points and mobile nodes

58

802.11 infrastructure mode• Mobile node sniffs based on a “Listen Interval”

– Listen Interval is multiple of the “beacon period”• Beacon period: typically 100ms

• During a Listen Interval– Access point

• buffers data for mobile node• sends out a traffic indication map (TIM), announcing buffered

data, every beacon period– Mobile node stays in power-saving mode

• After a Listen Interval– Mobile node checks TIM to see whether it gets buffered

data– If so, send “PS-Poll” asking for data

59

Buffering/sniffing in 802.11

Gast, 802.11 Wireless Network: The Definitive Guide

802.15.1/Bluetooth uses similar power-saving protocols: Hold and Sniff modes

60

Power profile of 802.11b

0

200

400

600

800

1000

1200

1400

1600

1 401 801 1201

Pow

er (m

W)

Time (1/4000 s)

61

Network variations

10080604020

20406080

100

1 21 41 61 81 101 121 141 161 181 201 221 241 261 281 301

Sign

al s

tren

gth

(%)

Distance (4-5 meters)

10080604020

20406080

100

1 21 41 61 81 101 121 141 161 181 201 221

Sign

al s

tren

gth

(%)

Distance (2-3 meters)

Cellular

Wi-Fi

Cellular

Wi-Fi

10080604020

20406080

100

1 121 241 361 481 601 721 841 961 1081 1201 1321 1441 1561 1681 1801 1921 2041 2161 2281 2401

Sign

al s

tren

gth

(%)

Time (minutes)

Cellular

Wi-Fi