Академический Документы
Профессиональный Документы
Культура Документы
• Logic level
• logic gates / logic equations
• nets / bits
• Physical level
• transistors / wires
• polygons
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
• Material properties
• mobility (Si) -- µn = 1250 cm2 / V sec & µp = 480 cm2 / V sec & R ~ µ-1
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Material properties
Mask levels 12 to 16 12 to 20 6 to 10
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
• Mobility -- gate
µn = 1250 cm2 / V sec & µp = 480 cm2 / V sec drain
w
Rp||Rp ~ Rn+Rn
Rondown CL
~CL· Ronup ~CL· Rondown Rp ~ 4·Rn (wp=wn)
CMOS DCFL
VDD
LPw
LO
input ξ output
LGw CL RL
PDC = 0 i1 = IDD or i2 = IDD
PAC = CV2f PDC = PAC = VDDIDD LPCB
CMOS
DCFL
A B
Vth
Vaa’
A’ B’ Vbb’
frequency
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Process steps
• Mask 1 - deep p-well diffusions nMOS transistors
• Mask 2 - thinox regions
• Mask 3 - polysilicon for “gate wires”
• Mask 4 - p-diffusion pMOS transistors, p +-mask & mask 2
• Mask 5 - n-diffusion nMOS transistors, inverted p +-mask & mask 2
• Mask 6 - contact cuts
• Mask 7 - metal layer
• Mask 8 - overglass layer overall passivation, access to bonding pads
7
6
5
4
3
2
1
V DD V in Vout V SS
p
n
VDD V in V o ut V SS
n -w e ll
n
p
VDD V in V ou t V SS
tw in -tu b
e p ita x ia l
n p la y e r
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Layout rules
• Wire width - wi
• w1>2, wi>wi-1 wiλ siλ
• Wire separation - si
• s1=3, si≥si-1
Layout environments
• Cell generation
• Programmable logic arrays (PLA)
• Transistor chaining
• Weinberger arrays & gate matrices
• Layout environments
• Standard cells
• Gate arrays Transistor chaining
• Sea-of-gates
• Field-Programmable Gate Arrays (FPGA)
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Layout methodologies
• Partitioning
• Floorplanning
• initial placement
• Placement
• fixed modules
• Global routing
• Detailed routing
• Layout optimization
• Layout verification
Partitioning
• Constructive approaches B D
• hierarchical clustering
• Iterative improvements
3
• Kernighan-Lin heuristic
2
cut-lines
• Weights
1
• module size
• number of connections A B C D
• number of I/O-s
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Floorplanning
• Sliceable floorplan
• two slices
• templates
• Hierarchical approaches
• bottom-up approach
• top-down approach
• Soft-computational approaches
• simulated annealing
• genetic algorithms
Placement
4
• Improvement of the initial floorplan
1 2
3
• Constructive heuristics 3
2
4
• Iterative heuristics 1
• Soft-computing 1
4
2
3
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Routing
4 3 2 3 4 5
• Maze running 9
9
8 9
9
9 8 9
3
2
2
1
1 2 3 4
1 2 3
5
4 5
9 8 7 8 9 8 7 8 9 3 2 1 2 3 4 5
• memory usage! 9 8 7 6 6 7 8 4 3 2 3 4
9 6 5 4 3 4 5 6 7 5 4 4 5 4 3 4 5
• bidirectional search 8 9 5 4 3 2 3 7 8 5 5 4 3 2 3
7 3 2 1 2 6 7 3 2 1 2
• minimum cost paths 6 5 4 3
7 6 5 4
2
3
1 1
2 1 2 3
5
4
6
5
5 4 3 2 1
5 4 3 2
1 5
1 2 3 4 5
8 7 6 5 4 3 2 3 4 5 6 5 4 3 2 3 4 5
• multilayer routing
• multiterminal routing
• Line searching
• track graph
Routing
• Global routing
• dividing routing task into smaller sub-tasks
• routing channels
• Detailed routing
• routing inside channels
• Layout optimization
• channel compaction
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Delay modelling
w
• Delays in wires can not be neglected anymore!
l
h
• Cw = fr εox w l / tox Rw = ρ l / w h tox
R/2 R/2 R
C C/2 C/2
• Elmore delay - TELM = ½RwCw + RwCL
T-model Π-model
Delay modelling
w1_3 w1_4 n2
n2
w1_8
n1 n1 w1_7
w1 w1_2
n3
n3
w1_1
w1_5 w1_6
D e p a r t m e n t o f C o m p u t e r E n g i n e e r i n g
Packaging
Packaging hierarchy
Crystal
1st level packaging
Multi-crystal-module / Chip
2nd level packaging
PCB
3rd level packaging
Mother-board
4th level packaging
Rack
Technology
requirements Higher Higher Lower Higher Higher
Packaging
constraints Lower Lower Higher Lower Lower