jan m. rabaey anantha chandrakasan borivoje nikolicbaccaran/rabaey/chapter01.pdf · jan m. rabaey....

EE1411

© Digital Integrated Circuits2nd Introduction

Digital Integrated Digital Integrated CircuitsCircuitsA Design PerspectiveA Design PerspectiveJan M. RabaeyAnantha ChandrakasanBorivoje Nikolic

IntroductionIntroductionJuly 30, 2002

EE1412


What is this book all about?What is this book all about?Introduction to digital integrated circuits.

CMOS devices and manufacturing technology. CMOS inverters and gates. Propagation delay, noise margins, and power dissipation. Sequential circuits. Arithmetic, interconnect, and memories. Programmable logic arrays. Design methodologies.

What will you learn?Understanding, designing, and optimizing digital circuits with respect to different quality metrics: cost, speed, power dissipation, and reliability

EE1413


Digital Integrated CircuitsDigital Integrated CircuitsIntroduction: Issues in digital designThe CMOS inverterCombinational logic structuresSequential logic gatesDesign methodologiesInterconnect: R, L and CTimingArithmetic building blocksMemories and array structures

EE1414


IntroductionIntroduction

Why is designing digital ICs different today than it was before?Will it change in future?

EE1415


The First ComputerThe First Computer

The BabbageDifference Engine(1832)25,000 partscost: £17,470

EE1416


ENIAC ENIAC -- The first electronic computer (1946)The first electronic computer (1946)

EE1417


The Transistor RevolutionThe Transistor Revolution

First transistorBell Labs, 1948

EE1418


The First Integrated Circuits The First Integrated Circuits

Bipolar logic1960’s

ECL 3-input GateMotorola 1966

EE1419


Intel 4004 MicroIntel 4004 Micro--ProcessorProcessor

19711000 transistors1 MHz operation

EE14110


Intel Pentium (IV) microprocessorIntel Pentium (IV) microprocessor

EE14111


Moore’s LawMoore’s Law

In 1965, Gordon Moore noted that the number of transistors on a chip doubled every 18 to 24 months.

He made a prediction that semiconductor technology will double its effectiveness every 18 months

EE14112


Moore’s LawMoore’s Law16151413121110

9876543210

1959

1960

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1975

LOG

2 OF

THE

NU

MB

ER O

FC

OM

PON

ENTS

PER

INTE

GR

ATE

D F

UN

CTI

ON

Electronics, April 19, 1965.

EE14113


Evolution in ComplexityEvolution in Complexity

EE14114


Transistor CountsTransistor Counts

808680286

i386i486

Pentium®Pentium® Pro

1 Billion 1 Billion TransistorsTransistors

Source: IntelSource: Intel

Pentium® IIPentium® III

K1,000,000

100,000

10,000

1,000

100

10

11975 1980 1985 1990 1995 2000 2005 2010

ProjectedProjected

Courtesy, Intel

EE14115


Moore’s law in MicroprocessorsMoore’s law in Microprocessors

400480088080

8085 8086286

386486 Pentium® proc

P6

0.001

0.01

0.1

1

10

100

1000Tr

ansi

stor

s (M

T)

2X growth in 1.96 years!

Transistors on Lead Microprocessors double every 2 yearsTransistors on Lead Microprocessors double every 2 years

1970 1980 1990 2000 2010Year

Courtesy, Intel

EE14116


Die Size GrowthDie Size Growth

40048008

80808085

8086286

386486 Pentium ® procP6

10

Die

siz

e (m

m)

~7% growth per year~2X growth in 10 years

100

11970 1980 1990 2000 2010

Year

Die size grows by 14% to satisfy Moore’s LawDie size grows by 14% to satisfy Moore’s Law

Courtesy, Intel

EE14117


FrequencyFrequency

P6Pentium ® proc

48638628680868085

808080084004

Freq

uenc

y (M

hz)

Doubles every2 years

10000

1000

100

10

1

0.11970 1980 1990 2000 2010

Year

Lead Microprocessors frequency doubles every 2 yearsLead Microprocessors frequency doubles every 2 years

Courtesy, Intel

EE14118


Power DissipationPower DissipationP6

Pentium ® proc

486386

2868086

808580808008

4004

Pow

er (W

atts

)100

10

1

0.11971 1974 1978 1985 1992 2000

Year

Lead Microprocessors power continues to increaseLead Microprocessors power continues to increase

Courtesy, Intel

EE14119


Power will be a major problemPower will be a major problem100000

5KW 18KW

1.5KW 500W

4004800880808085

8086286

386486

Pentium® proc

10000

1

10

100

Pow

er (W

atts

)

1000

0.11971 1974 1978 1985 1992 2000 2004 2008

Year

Power delivery and dissipation will be prohibitivePower delivery and dissipation will be prohibitive

Courtesy, Intel

EE14120


Power densityPower density

400480088080

8085

8086

286 386486

Pentium® procP6

1

10

100

1000

10000

1970 1980 1990 2000 2010Year

Pow

er D

ensi

ty (W

/cm

2)

Hot Plate

NuclearReactor

RocketNozzle

Power density too high to keep junctions at low tempPower density too high to keep junctions at low temp

Courtesy, Intel

EE14121


Not Only MicroprocessorsNot Only Microprocessors

Analog Baseband

Digital Baseband(DSP + MCU)

PowerManagement

Small Signal RF

PowerRF

CellPhone

Digital Cellular Market(Phones Shipped)

1996 1997 1998 1999 2000

Units 48M 86M 162M 260M 435M

(data from Texas Instruments)(data from Texas Instruments)

EE14122


Challenges in Digital DesignChallenges in Digital Design

∝ 1/DSM

“Microscopic Problems”• Ultra-high speed design• Interconnect• Noise, Crosstalk• Reliability, Manufacturability• Power Dissipation• Clock distribution.

Everything Looks a Little Different

∝ DSM

?

“Macroscopic Issues”• Time-to-Market• Millions of Gates• High-Level Abstractions• Reuse & IP: Portability• Predictability• etc.

…and There’s a Lot of Them!

EE14123


Productivity TrendsProductivity Trends

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

2003

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

2001

2005

2007

2009

10

100

1,000

10,000

100,000

1,000,000

10,000,000

100,000,000Logic Tr./ChipTr./Staff Month.

xxxx

xx

x21%/Yr. compound

Productivity growth rate

x

58%/Yr. compoundedComplexity growth rate

10,000

1,000

100

10

1

0.1

0.01

0.001

Logi

c Tr

ansi

stor

per

Chi

p(M

)

0.01

0.1

1

10

100

1,000

10,000

100,000

Prod

uctiv

ity(K

) Tra

ns./S

taff

-Mo.

Com

plex

ity

Source: Sematech

Complexity outpaces design productivity

Courtesy, ITRS Roadmap

EE14124


Why Scaling?Why Scaling?Technology shrinks by 0.7/generationWith every generation can integrate 2x more functions per chip; chip cost does not increase significantlyCost of a function decreases by 2xBut …

How to design chips with more and more functions?Design engineering population does not double every two years…

Hence, a need for more efficient design methodsExploit different levels of abstraction

EE14125


Design Abstraction LevelsDesign Abstraction Levels

n+n+S

GD

+

DEVICE

CIRCUIT

GATE

MODULE

SYSTEM

EE14126


Design MetricsDesign Metrics

How to evaluate performance of a digital circuit (gate, block, …)?

CostReliabilityScalabilitySpeed (delay, operating frequency) Power dissipationEnergy to perform a function

EE14127


Cost of Integrated CircuitsCost of Integrated Circuits

NRE (non-recurrent engineering) costsdesign time and effort, mask generationone-time cost factor

Recurrent costssilicon processing, packaging, testproportional to volumeproportional to chip area

EE14128


NRE Cost is IncreasingNRE Cost is Increasing

EE14129


Die CostDie Cost

Single die

Wafer

Going up to 12” (30cm)

From http://www.amd.com

EE14130


Cost per TransistorCost per Transistor

Fabrication capital cost per transistor (Moore’s law)

0.00000010.0000001

0.0000010.000001

0.000010.00001

0.00010.0001

0.0010.001

0.010.01

0.10.111

cost: cost: ¢¢--perper--transistortransistor

19821982 19851985 19881988 19911991 19941994 19971997 20002000 20032003 20062006 20092009 20122012

EE14131


YieldYield%100

per wafer chips ofnumber Totalper wafer chips good of No.

×=Y

yield Dieper wafer DiescostWafer cost Die×

=

( )area die2

diameterwafer area die

diameter/2wafer per wafer Dies2

××π

−×π

=

EE14132


DefectsDefects

α−

α×

+=area dieareaunit per defects1yield die

α is approximately 3

4area) (die cost die f=

EE14133


Some Examples (1994)Some Examples (1994)

$4179%402961.5$15000.803Pentium

$27213%482561.6$17000.703Super Sparc

$14919%532341.2$15000.703DEC Alpha

$7327%661961.0$13000.803HP PA 7100

$5328%1151211.3$17000.804Power PC 601

$1254%181811.0$12000.803486 DX2

$471%360431.0$9000.902386DX

Die cost

YieldDies/wafer

Area mm2

Def./ cm2

Wafer cost

Line width

Metal layers

Chip

EE14134


ReliabilityReliability――Noise in Digital Integrated CircuitsNoise in Digital Integrated Circuits

VDDv(t)

i(t)

(a) Inductive coupling (b) Capacitive coupling (c) Power and ground

noise

EE14135


DC OperationDC OperationVoltage Transfer CharacteristicVoltage Transfer Characteristic

V(x)

V(y)

VOH

VOL

VM

VOHVOL

fV(y)=V(x)

Switching Threshold

Nominal Voltage Levels

VOH = f(VOL)VOL = f(VOH)VM = f(VM)

EE14136


Mapping between analog and digital signalsMapping between analog and digital signals

"1"

"0"

VOHVIH

VIL VOL

UndefinedRegion

V(x)

V(y)

VOH

VOL

VIHV

IL

Slope = -1

Slope = -1

EE14137


Definition of Noise MarginsDefinition of Noise Margins

NMH

"1"

VOH Noise margin highVIHUndefined

RegionVILNML Noise margin lowVOL

"0"

Gate InputGate Output

EE14138


Noise BudgetNoise Budget

Allocates gross noise margin to expected sources of noiseSources: supply noise, cross talk, interference, offsetDifferentiate between fixed and proportional noise sources

EE14139


Key Reliability PropertiesKey Reliability PropertiesAbsolute noise margin values are deceptive

a floating node is more easily disturbed than a node driven by a low impedance (in terms of voltage)

Noise immunity is the more important metric –the capability to suppress noise sourcesKey metrics: Noise transfer functions, Output

impedance of the driver and input impedance of the receiver;

EE14140


Regenerative PropertyRegenerative Property

in

out out

in

f(v)

finv(v)

finv (v)

f(v)

v3

v1

v2 v0v0 v2

v3

v1

Non-RegenerativeRegenerative

EE14141


Regenerative PropertyRegenerative Property

(a) A chain of inverters

v0 v1 v2 v3 v4v5 v6

…

0 2 4 6 8 10t (nsec)

–1

1

3

5

V (V

olt)

v0

v1 v2

(b) Simulated response of chain of MOS inverters

EE14142


FanFan--in and Fanin and Fan--outout

N

M

(a) Fan-out N

(b) Fan-in M

EE14143


The Ideal GateThe Ideal Gate

Vin

Vout

g=−∞

Ri = ∞

Ro = 0

Fanout = ∞

NMH = NML = VDD/2

EE14144


An OldAn Old--time Invertertime Inverter

0.0 1.0 2.0 3.0 4.0 5.0Vin (V)

1.0

2.0

3.0

4.0

5.0V

out (

V)

VMNMH

NML

EE14145


Delay DefinitionsDelay Definitions

Vout

t

Vin

50%

tpHL tpLH90%

50%

10% t

trtf

EE14146


Ring OscillatorRing Oscillator

v0 v1 v2 v3 v4 v5

v0 v1 v5

T = 2 × tp × N

EE14147


A FirstA First--Order RC NetworkOrder RC Network

vout

vin C

R

tp = ln (2) τ = 0.69 RC

Important model – matches delay of inverter

EE14148


Power DissipationPower Dissipation

Instantaneous power: p(t) = v(t)i(t) = Vsupplyi(t)

Peak power: Ppeak = Vsupplyipeak

Average power:

( )∫ ∫+ +

==Tt

tTt

t supplysupply

ave dttiT

Vdttp

TP )(1

EE14149


Energy and EnergyEnergy and Energy--DelayDelay

Power-Delay Product (PDP) =

E = Energy per operation = Pav × tp

Energy-Delay Product (EDP) =

quality metric of gate = E × tp

EE14150


A FirstA First--Order RC NetworkOrder RC Network

voutR

vin CL

E0 1→ P t( )dt

0

T

∫ Vdd isupply t( )dt

0

T

∫ Vdd CLdVout0

Vdd

∫ CL Vdd• 2= = = =

Ecap Pcap t( )dt

0

T

∫ Vouticap t( )dt

0

T

∫ CLVoutdVout0

Vdd

∫12---C

LVdd• 2= = = =

EE14151


SummarySummaryDigital integrated circuits have come a long way and still have quite some potential left for the coming decadesSome interesting challenges ahead

Getting a clear perspective on the challenges and potential solutions is the purpose of this course

Understanding the design metrics that govern digital design is crucial

Cost, reliability, speed, power and energy dissipation

jan m. rabaey anantha chandrakasan borivoje nikolicbaccaran/rabaey/chapter01.pdf · jan m. rabaey....

Documents