CMOS VLSI Design

advertisement
CMOS VLSI Design
Digital Design
Digital Design
CMOS VLSI
Slide 1
Overview






Physical principles
Combinational logic
Sequential logic
Datapath
Memories
Trends
Digital Design
CMOS VLSI
Slide 2
Dopants





Silicon is a semiconductor
Pure silicon has no free carriers and conducts poorly
Adding dopants increases the conductivity
Group V: extra electron (n-type)
Group III: missing electron, called hole (p-type)
Digital Design
Si
Si
Si
Si
Si
Si
As
Si
Si
B
Si
Si
Si
Si
Si
-
+
CMOS VLSI
+
-
Si
Si
Si
Slide 3
nMOS Operation
 Body is commonly tied to ground (0 V)
 When the gate is at a low voltage:
– P-type body is at low voltage
– Source-body and drain-body diodes are OFF
– No current flows, transistor is OFF
Source
Gate
Drain
Polysilicon
SiO2
0
n+
n+
S
p
Digital Design
D
bulk Si
CMOS VLSI
Slide 4
Transistors as Switches
 We can view MOS transistors as electrically
controlled switches
 Voltage at gate controls path from source to drain
d
nMOS
pMOS
g=1
d
d
OFF
g
ON
s
s
s
d
d
d
g
OFF
ON
s
Digital Design
g=0
s
CMOS VLSI
s
Slide 5
CMOS Inverter
A
VDD
Y
0
1
A
A
Y
Y
GND
Digital Design
CMOS VLSI
Slide 6
Inverter Cross-section
 Typically use p-type substrate for nMOS transistors
 Requires n-well for body of pMOS transistors
A
GND
VDD
Y
SiO2
n+ diffusion
n+
n+
p+
p+
n well
p substrate
nMOS transistor
Digital Design
p+ diffusion
polysilicon
metal1
pMOS transistor
CMOS VLSI
Slide 7
Inverter Mask Set
 Transistors and wires are defined by masks
 Cross-section taken along dashed line
A
Y
GND
VDD
nMOS transistor
well tap
substrate tap
Digital Design
pMOS transistor
CMOS VLSI
Slide 8
Fabrication Steps
 Start with blank wafer
 Build inverter from the bottom up
 First step will be to form the n-well
– Cover wafer with protective layer of SiO2 (oxide)
– Remove layer where n-well should be built
– Implant or diffuse n dopants into exposed wafer
– Strip off SiO2
p substrate
Digital Design
CMOS VLSI
Slide 9
Oxidation
 Grow SiO2 on top of Si wafer
– 900 – 1200 C with H2O or O2 in oxidation furnace
SiO2
p substrate
Digital Design
CMOS VLSI
Slide 10
Photoresist
 Spin on photoresist
– Photoresist is a light-sensitive organic polymer
– Softens where exposed to light
Photoresist
SiO2
p substrate
Digital Design
CMOS VLSI
Slide 11
Lithography
 Expose photoresist through n-well mask
 Strip off exposed photoresist
Photoresist
SiO2
p substrate
Digital Design
CMOS VLSI
Slide 12
Etch
 Etch oxide with hydrofluoric acid (HF)
– Seeps through skin and eats bone; nasty stuff!!!
 Only attacks oxide where resist has been exposed
Photoresist
SiO2
p substrate
Digital Design
CMOS VLSI
Slide 13
Strip Photoresist
 Strip off remaining photoresist
– Use mixture of acids called piranah etch
 Necessary so resist doesn’t melt in next step
SiO2
p substrate
Digital Design
CMOS VLSI
Slide 14
n-well
 n-well is formed with diffusion or ion implantation
 Diffusion
– Place wafer in furnace with arsenic gas
– Heat until As atoms diffuse into exposed Si
 Ion Implanatation
– Blast wafer with beam of As ions
– Ions blocked by SiO2, only enter exposed Si
SiO2
n well
Digital Design
CMOS VLSI
Slide 15
Simplified Design Rules
 Conservative rules to get you started
Digital Design
CMOS VLSI
Slide 16
Complementary CMOS
 Complementary CMOS logic gates
– nMOS pull-down network
– pMOS pull-up network
inputs
– a.k.a. static CMOS
Pull-up OFF
Pull-up ON
Pull-down OFF Z (float)
1
Pull-down ON
X (crowbar)
Digital Design
0
CMOS VLSI
pMOS
pull-up
network
output
nMOS
pull-down
network
Slide 17
Example: NAND3





Horizontal N-diffusion and p-diffusion strips
Vertical polysilicon gates
Metal1 VDD rail at top
Metal1 GND rail at bottom
32 l by 40 l
Digital Design
CMOS VLSI
Slide 18
I-V Characteristics
 In Linear region, Ids depends on
– How much charge is in the channel?
– How fast is the charge moving?
Digital Design
CMOS VLSI
Slide 19
Channel Charge
 MOS structure looks like parallel plate capacitor
while operating in inversion
– Gate – oxide – channel
 Qchannel = CV
Cox = eox / tox
 C = Cg = eoxWL/tox = CoxWL
 V = Vgc – Vt = (Vgs – Vds/2) – Vt
gate
Vg
polysilicon
gate
W
tox
n+
L
n+
SiO2 gate oxide
(good insulator, eox = 3.9)
+
+
Cg Vgd drain
source Vgs
Vs
Vd
channel
+
n+
n+
Vds
p-type body
p-type body
Digital Design
CMOS VLSI
Slide 20
Carrier velocity
 Charge is carried by e Carrier velocity v proportional to lateral E-field
between source and drain
 v = mE
m called mobility
 E = Vds/L
 Time for carrier to cross channel:
– t=L/v
Digital Design
CMOS VLSI
Slide 21
nMOS Linear I-V
 Now we know
– How much charge Qchannel is in the channel
– How much time t each carrier takes to cross
Qchannel
I ds 
t
W
 mCox
L
V  V  Vds
 gs t
2

V
  Vgs  Vt  ds Vds
2

Digital Design
CMOS VLSI
V
 ds

W
 = mCox
L
Slide 22
Example
 Example: a 0.6 mm process from AMI semiconductor
– tox = 100 Å
– m = 350 cm2/V*s
2.5
V =5
– Vt = 0.7 V
2
 Plot Ids vs. Vds
1.5
V =4
– Vgs = 0, 1, 2, 3, 4, 5
1
V =3
– Use W/L = 4/2 l
0.5
Ids (mA)
gs
gs
gs
0
Vgs = 2
Vgs = 1
0
 3.9  8.85  1014   W 
W
W
  mCox   350 

120
m A /V 2
 

8
L
L
 100  10
 L 
Digital Design
CMOS VLSI
1
2
3
4
5
Vds
Slide 23
Capacitance
 Any two conductors separated by an insulator have
capacitance
 Gate to channel capacitor is very important
– Creates channel charge necessary for operation
 Source and drain have capacitance to body
– Across reverse-biased diodes
– Called diffusion capacitance because it is
associated with source/drain diffusion
Digital Design
CMOS VLSI
Slide 24
Gate Capacitance
 Approximate channel as connected to source
 Cgs = eoxWL/tox = CoxWL = CpermicronW
 Cpermicron is typically about 2 fF/mm
polysilicon
gate
W
tox
n+
L
n+
SiO2 gate oxide
(good insulator, eox = 3.9e0)
p-type body
Digital Design
CMOS VLSI
Slide 25
Diffusion Capacitance
 Csb, Cdb
 Undesirable, called parasitic capacitance
 Capacitance depends on area and perimeter
– Use small diffusion nodes
– Comparable to Cg
for contacted diff
– ½ Cg for uncontacted
– Varies with process
Digital Design
CMOS VLSI
Slide 26
RC Delay Model
 Use equivalent circuits for MOS transistors
– Ideal switch + capacitance and ON resistance
– Unit nMOS has resistance R, capacitance C
– Unit pMOS has resistance 2R, capacitance C
 Capacitance proportional to width
 Resistance inversely proportional to width
d
g
d
k
s
s
kC
R/k
2R/k
g
g
kC
kC
s
Digital Design
kC
d
k
s
kC
g
kC
d
CMOS VLSI
Slide 27
Introduction
 Chips are mostly made of wires called interconnect
– In stick diagram, wires set size
– Transistors are little things under the wires
– Many layers of wires
 Wires are as important as transistors
– Speed
– Power
– Noise
 Alternating layers run orthogonally
Digital Design
CMOS VLSI
Slide 28
Wire Capacitance
 Wire has capacitance per unit length
– To neighbors
– To layers above and below
 Ctotal = Ctop + Cbot + 2Cadj
s
w
layer n+1
h2
Ctop
t
h1
layer n
Cbot
Cadj
layer n-1
Digital Design
CMOS VLSI
Slide 29
Lumped Element Models
 Wires are a distributed system
– Approximate with lumped element models
N segments
R
R/N
C
R/N
C/N
C/N
R
R
C
L-model
C/2
R/N
R/N
C/N
C/N
R/2 R/2
C/2
p-model
C
T-model
 3-segment p-model is accurate to 3% in simulation
 L-model needs 100 segments for same accuracy!
 Use single segment p-model for Elmore delay
Digital Design
CMOS VLSI
Slide 30
Crosstalk
 A capacitor does not like to change its voltage
instantaneously.
 A wire has high capacitance to its neighbor.
– When the neighbor switches from 1-> 0 or 0->1,
the wire tends to switch too.
– Called capacitive coupling or crosstalk.
 Crosstalk effects
– Noise on nonswitching wires
– Increased delay on switching wires
Digital Design
CMOS VLSI
Slide 31
Coupling Waveforms
 Simulated coupling for Cadj = Cvictim
Aggressor
1.8
1.5
1.2
Victim (undriven): 50%
0.9
0.6
Victim (half size driver): 16%
Victim (equal size driver): 8%
0.3
Victim (double size driver): 4%
0
0
200
400
600
800
1000
1200
1400
1800
2000
t(ps)
Digital Design
CMOS VLSI
Slide 32
Introduction
 What makes a circuit fast?
– I = C dV/dt -> tpd  (C/I) DV
– low capacitance
– high current
4
B
– small swing
4
A
 Logical effort is proportional to C/I
1
1
 pMOS are the enemy!
– High capacitance for a given current
 Can we take the pMOS capacitance off the input?
 Various circuit families try to do this…
Digital Design
CMOS VLSI
Y
Slide 33
Pseudo-nMOS
 In the old days, nMOS processes had no pMOS
– Instead, use pull-up transistor that is always ON
 In CMOS, use a pMOS that is always ON
– Ratio issue
– Make pMOS about ¼ effective strength of
pulldown network
1.8
1.5
load
P/2
1.2
P = 24
Ids
Vout 0.9
Vout
0.6
P = 14
16/2
0.3
Vin
P=4
0
0
0.3
0.6
0.9
1.2
1.5
1.8
Vin
Digital Design
CMOS VLSI
Slide 34
Dynamic Logic
 Dynamic gates uses a clocked pMOS pullup
 Two modes: precharge and evaluate
2
A

2/3
Y
1
Y
1
A
Static
4/3
Pseudo-nMOS

Precharge
Y
A
1
Dynamic
Evaluate
Precharge
Y
Digital Design
CMOS VLSI
Slide 35
Pass Transistor Circuits
 Use pass transistors like switches to do logic
 Inputs drive diffusion terminals as well as gates
 CMOS + Transmission Gates:
– 2-input multiplexer
– Gates should be restoring
S
S
A
A
S
S
Y
B
B
S
S
Digital Design
Y
CMOS VLSI
Slide 36
Sequencing
 Combinational logic
– output depends on current inputs
 Sequential logic
– output depends on current and previous inputs
– Requires separating previous, current, future
– Called state or tokens
– Ex: FSM, pipeline
clk
in
clk
clk
clk
out
CL
CL
Finite State Machine
Digital Design
CL
Pipeline
CMOS VLSI
Slide 37
Sequencing Overhead
 Use flip-flops to delay fast tokens so they move
through exactly one stage each cycle.
 Inevitably adds some delay to the slow tokens
 Makes circuit slower than just the logic delay
– Called sequencing overhead
 Some people call this clocking overhead
– But it applies to asynchronous circuits too
– Inevitable side effect of maintaining sequence
Digital Design
CMOS VLSI
Slide 38
Sequencing Elements
 Latch: Level sensitive
– a.k.a. transparent latch, D latch
 Flip-flop: edge triggered
– A.k.a. master-slave flip-flop, D flip-flop, D register
 Timing Diagrams
– Transparent
– Opaque
– Edge-trigger
clk
Q
D
Flop
D
Latch
clk
Q
clk
D
Q (latch)
Q (flop)
Digital Design
CMOS VLSI
Slide 39
Latch Design
 Buffered output
+ No backdriving

X
D

 Widely used in standard cells
+ Very robust (most important)
- Rather large
- Rather slow (1.5 – 2 FO4 delays)
- High clock loading
Digital Design
CMOS VLSI
Q


Slide 40
Sequencing Methods
clk
clk
Combinational Logic
tnonoverlap
Combinational
Logic
Combinational
Logic
Half-Cycle 1
tpw
p
Combinational Logic
Latch
p
Latch
Pulsed Latches
CMOS VLSI
1
Latch
2
Latch
1
p
tnonoverlap
Tc/2
2
Latch
2-Phase Transparent Latches
1
Half-Cycle 1
Digital Design
Flop
clk
Flop
Flip-Flops
 Flip-flops
 2-Phase Latches
 Pulsed Latches
Tc
Slide 41
Summary
 Flip-Flops:
– Very easy to use, supported by all tools
 2-Phase Transparent Latches:
– Lots of skew tolerance and time borrowing
 Pulsed Latches:
– Fast, some skew tol & borrow, hold time risk
Digital Design
CMOS VLSI
Slide 42
Full Adder Design I
 Brute force implementation from eqns
S  A B C
Cout  MAJ ( A, B, C )
A
A
A
S
C
A
B
B
C
C
MAJ
C
B
C
B
B
A
A
Digital Design
C
B
S
Cout
B
A
B
A
B
C
A
B
C
B
B
C
A
C
A
B
Cout
B
A
CMOS VLSI
Slide 43
Carry-Skip Adder
 Carry-ripple is slow through all N stages
 Carry-skip allows carry to skip over groups of n bits
– Decision based on n-bit propagate signal
Cout
A16:13 B16:13
A12:9 B12:9
A8:5 B8:5
A4:1
P16:13
P12:9
P8:5
P4:1
1
0
C12
+
S16:13
Digital Design
1
0
C8
+
S12:9
CMOS VLSI
1
0
C4
+
S8:5
B4:1
1
0
+
Cin
S4:1
Slide 44
Tree Adder
 If lookahead is good, lookahead across lookahead!
– Recursive lookahead gives O(log N) delay
 Many variations on tree adders
Digital Design
CMOS VLSI
Slide 45
Memory Arrays
Memory Arrays
Random Access Memory
Read/Write Memory
(RAM)
(Volatile)
Static RAM
(SRAM)
Dynamic RAM
(DRAM)
Mask ROM
Programmable
ROM
(PROM)
Digital Design
Content Addressable Memory
(CAM)
Serial Access Memory
Read Only Memory
(ROM)
(Nonvolatile)
Shift Registers
Serial In
Parallel Out
(SIPO)
Erasable
Programmable
ROM
(EPROM)
CMOS VLSI
Queues
Parallel In
Serial Out
(PISO)
Electrically
Erasable
Programmable
ROM
(EEPROM)
First In
First Out
(FIFO)
Last In
First Out
(LIFO)
Flash ROM
Slide 46
Array Architecture
 2n words of 2m bits each
 If n >> m, fold by 2k into fewer rows of more columns
wordlines
bitline conditioning
bitlines
row decoder
memory cells:
2n-k rows x
2m+k columns
n-k
column
circuitry
k
n
column
decoder
2m bits
 Good regularity – easy to design
 Very high density if good cells are used
Digital Design
CMOS VLSI
Slide 47
6T SRAM Cell
 Cell size accounts for most of array size
– Reduce cell size at expense of complexity
 6T SRAM Cell
– Used in most commercial chips
– Data stored in cross-coupled inverters
 Read:
bit
– Precharge bit, bit_b
word
– Raise wordline
 Write:
– Drive data onto bit, bit_b
– Raise wordline
Digital Design
CMOS VLSI
bit_b
Slide 48
SRAM Sizing
 High bitlines must not overpower inverters during
reads
 But low bitlines must write new value into cell
bit_b
bit
word
weak
med
med
A
A_b
strong
Digital Design
CMOS VLSI
Slide 49
Decoders
 n:2n decoder consists of 2n n-input AND gates
– One needed for each row of memory
– Build AND from NAND or NOR gates
Static CMOS
A1
Pseudo-nMOS
A0
A1
word0
word1
Digital Design
1
1
8
A1
1
4
A0
1
A0
word0
word
word1
word2
word2
word3
word3
CMOS VLSI
A0
1/2
4
16
A1
1
1
2
8
word
Slide 50
Decoder Layout
 Decoders must be pitch-matched to SRAM cell
– Requires very skinny gates
A3
A3
A2
A2
A1
A1
A0
A0
VDD
word
GND
buffer inverter
NAND gate
Digital Design
CMOS VLSI
Slide 51
Sense Amplifiers
 Bitlines have many cells attached
– Ex: 32-kbit SRAM has 256 rows x 128 cols
– 128 cells on each bitline
 tpd  (C/I) DV
– Even with shared diffusion contacts, 64C of
diffusion capacitance (big C)
– Discharged slowly through small transistors
(small I)
 Sense amplifiers are triggered on small voltage
swing (reduce DV)
Digital Design
CMOS VLSI
Slide 52
Queues
 Queues allow data to be read and written at different
rates.
 Read and write each use their own clock, data
 Queue indicates whether it is full or empty
 Build with SRAM and read/write counters (pointers)
WriteClk
WriteData
ReadClk
Queue
FULL
Digital Design
ReadData
EMPTY
CMOS VLSI
Slide 53
CAMs
 Extension of ordinary memory (e.g. SRAM)
– Read and write memory as usual
– Also match to see which words contain a key
adr
data/key
read
CAM
match
write
Digital Design
CMOS VLSI
Slide 54
10T CAM Cell
 Add four match transistors to 6T SRAM
– 56 x 43 l unit cell
bit
bit_b
word
cell_b
cell
match
Digital Design
CMOS VLSI
Slide 55
CAM Cell Operation
address
read/write
Digital Design
CMOS VLSI
CAM cell
clk
weak
miss
match0
row decoder
 Read and write like ordinary SRAM
 For matching:
– Leave wordline low
– Precharge matchlines
– Place key on bitlines
– Matchlines evaluate
 Miss line
– Pseudo-nMOS NOR of match lines
– Goes high if no words match
match1
match2
match3
column circuitry
data
Slide 56
ROM Example
 4-word x 6-bit ROM
– Represented with dot diagram
– Dots indicate 1’s in ROM
weak
pseudo-nMOS
pullups
A1 A0
Word 0: 010101
Word 1: 011001
Word 2: 100101
Word 3: 101010
2:4
DEC
ROM Array
Y5
Y4
Y3
Y2
Y1
Y0
Looks like 6 4-input pseudo-nMOS NORs
Digital Design
CMOS VLSI
Slide 57
PLAs
 A Programmable Logic Array performs any function
in sum-of-products form.
 Literals: inputs & complements
 Products / Minterms: AND of literals
 Outputs: OR of Minterms
bc
AND Plane
OR Plane
 Example: Full Adder
s  abc  abc  abc  abc
cout  ab  bc  ac
a
b
Inputs
Digital Design
CMOS VLSI
c
s
Minterms
ac
ab
abc
abc
abc
abc
cout
Outputs
Slide 58
PLA Schematic & Layout
AND Plane
OR Plane
bc
ac
ab
abc
abc
abc
abc
a
b
c
s
Digital Design
cout
CMOS VLSI
Slide 59
Ideal nMOS I-V Plot
 180 nm TSMC process
 Ideal Models
–  = 155(W/L) mA/V2
– Vt = 0.4 V
– VDD = 1.8 V
Ids (mA)
400
Vgs = 1.8
300
Vgs = 1.5
200
Vgs = 1.2
100
0
Digital Design
CMOS VLSI
Vgs = 0.9
Vgs = 0.6
0
0.3
0.6
0.9
1.2
1.5
1.8
Vds
Slide 60
Simulated nMOS I-V Plot
 180 nm TSMC process
 BSIM 3v3 SPICE models
I (mA)
 What differs?
250
– Less ON current
200
– No square law
150
– Current increases
100
in saturation
ds
Vgs = 1.8
Vgs = 1.5
Vgs = 1.2
Vgs = 0.9
50
Vgs = 0.6
0
0
0.3
0.6
0.9
1.2
1.5
Vds
Digital Design
CMOS VLSI
Slide 61
Velocity Saturation
 We assumed carrier velocity is proportional to E-field
– v = mElat = mVds/L
 At high fields, this ceases to be true
– Carriers scatter off atoms


– Velocity reaches vsat
• Electrons: 6-10 x 106 cm/s
 /2
• Holes: 4-8 x 106 cm/s
– Better model
slope = m
sat
sat
μElat
v
 vsat  μEsat
Elat
1
Esat
Digital Design
CMOS VLSI
0
0
Esat
2Esat
3Esat
Elat
Slide 62
Channel Length Modulation
 Reverse-biased p-n junctions form a depletion region
– Region between n and p with no carriers
– Width of depletion Ld region grows with reverse bias
V
V
GND
Source
Gate
Drain
– Leff = L – Ld
Depletion Region
Width: L
 Shorter Leff gives more current
– Ids increases with Vds
L
n+
n+
L
– Even in saturation
p GND bulk Si
DD
DD
d
eff
Digital Design
CMOS VLSI
Slide 63
Body Effect
 Vt: gate voltage necessary to invert channel
 Increases if source voltage increases because
source is connected to the channel
 Increase in Vt with Vs is called the body effect
Digital Design
CMOS VLSI
Slide 64
OFF Transistor Behavior
 What about current in cutoff?
I
 Simulated results
1 mA
 What differs?
Sub100 mA
threshold
– Current doesn’t go 10 mA Region
1 mA
to 0 in cutoff
100 nA
ds
10 nA
Saturation
Region
Vds = 1.8
Subthreshold
Slope
1 nA
100 pA
10 pA
Vt
0
0.3
0.6
0.9
1.2
1.5
1.8
Vgs
Digital Design
CMOS VLSI
Slide 65
Leakage Sources
 Subthreshold conduction
– Transistors can’t abruptly turn ON or OFF
 Junction leakage
– Reverse-biased PN junction diode current
 Gate leakage
– Tunneling through ultrathin gate dielectric
 Subthreshold leakage is the biggest source in
modern transistors
Digital Design
CMOS VLSI
Slide 66
Low Power Design
 Reduce dynamic power
– a: clock gating, sleep mode
– C: small transistors (esp. on clock), short wires
– VDD: lowest suitable voltage
– f: lowest suitable frequency
 Reduce static power
– Selectively use ratioed circuits
– Selectively use low Vt devices
– Leakage reduction:
stacked devices, body bias, low temperature
Digital Design
CMOS VLSI
Slide 67
Chip-to-Package Bonding
 Traditionally, chip is surrounded by pad frame
– Metal pads on 100 – 200 mm pitch
– Gold bond wires attach pads to package
– Lead frame distributes signals in package
– Metal heat spreader helps with cooling
Digital Design
CMOS VLSI
Slide 68
Bidirectional Pads
 Combine input and output pad
 Need tristate driver on output
– Use enable signal to set direction
– Optimized tristate avoids huge series transistors
PAD
En
Din
Dout
NAND
Dout
En
Y
Dout
NOR
Digital Design
CMOS VLSI
Slide 69
Device Scaling
Digital Design
CMOS VLSI
Slide 70
Interconnect Delay
Digital Design
CMOS VLSI
Slide 71
Download