Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Introduction
Digital Computer Arithmetic belongs to
Computer Architecture, however, it is also an
aspect of logic design.
The objective of Computer Arithmetic is to
develop appropriate algorithms that are
utilizing available hardware in the most
efficient way.
Ultimately, speed, power and chip area are
the most often used measures, making a
strong link between the algorithms and
technology of implementation.
Oklobdzija 2004
Computer Arithmetic
Basic Operations
Addition
Multiplication
Multiply-Add
Division
Evaluation of Functions
Multi-Media
Oklobdzija 2004
Computer Arithmetic
ai
Cout
bi
Full
Adder
Cin
si
si ai bi ci ai bi ci ai bi ci ai bi ci
ci 1 ai bi ci ai bi ci ai bi ci ai bi ci ai bi ai ci bi ci
Oklobdzija 2004
Computer Arithmetic
Outputs
ci
ai
bi
si
ci+1
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
0
1
1
0
1
0
0
1
0
0
0
1
0
1
1
1
Oklobdzija 2004
Computer Arithmetic
Propagate
Generate
Propagate
Generate
6
Full-Adder Implementation
Full Adder operations is defined by equations:
si ai bi ci ai bi ci ai bi ci ai bi ci ai bi ci pi ci
ci 1 ai bi ci ai bi ci ai bi g i pi ci
ai b
Carry-Propagate: pi ai bi
and Carry-Generate gi
g i a i bi
c out
c in
Computer Arithmetic
si
High-Speed Addition
ci 1 g i pi ci
a
g i ai bi
pi ai bi
c out
s
c in
si pi ci
si
Computer Arithmetic
Oklobdzija 2004
Computer Arithmetic
A1
B0
FA
S0
Co,0
(= C i,1)
A2
B1
FA
S1
C o,1
A3
B2
FA
Co,2
S2
FA
B3
Co,3
S3
Computer Arithmetic
10
Inversion Property
A
Ci
FA
Co
Ci
FA
Co
S
S A B C i = S A B C i
C o A B C i = Co A B Ci
From Rabaey
Oklobdzija 2004
Computer Arithmetic
11
A1
B0
FA
C o,0
S0
B1
FA
S1
A2
Co,1
A3
B2
FA
Odd Cell
C o,2
S2
B3
FA
C o,3
S3
Computer Arithmetic
12
Critical Path
c i+ 1
c out
ci
c in
Oklobdzija, ISCAS88
s i+ 2
Oklobdzija 2004
s i+ 1
Computer Arithmetic
si
13
Manchester Carry-Chain
Realization of the Carry Path
Simple and very popular scheme for implementation of
carry signal path
dd
dd
dd
dd
dd
dd
dd
dd
Generate
device
Carry out
Carry in
+
+ Propagate
device
Predischarge
& kill device
Oklobdzija 2004
Computer Arithmetic
14
Original Design
T. Kilburn, D. B. G. Edwards, D. Aspinall, "Parallel Addition in Digital Computers:
A New Fast "Carry" Circuit", Proceedings of IEE, Vol. 106, pt. B, p. 464, September 1959.
Oklobdzija 2004
Computer Arithmetic
15
VDD
Ci,0
P0
P1
P2
P3
P4
G0
G1
G2
G3
G4
Computer Arithmetic
16
C C
A
A
B
B
S
X O R /X N O R
U L T IP L E X E R
U FFER
A N D /N A N D
V
C C
A
A
B
B
V
A
A
B
B
C C
C C
U L T IP L E X E R
U FFER
O R /N O R
Oklobdzija 2004
Computer Arithmetic
17
Carry-Skip Adder
MacSorley, Proc IRE 1/61
Lehman, Burla, IRE Trans on Comp, 12/61
Oklobdzija 2004
Computer Arithmetic
18
Carry-Skip Adder
Ci,0
G1
FA
P0
C o,0
P0 G1
Ci,0
FA
FA
P0
C o,0
G1
P2
C o,1
G1
FA
FA
P2
Co,1
G2
P3
Co,2
G2
FA
FA
P3
C o,2
G3
G3
FA
Co,3
BP=P oP1 P2 P3
Multiplexer
P0
Co,3
Bypass
From Rabaey
Oklobdzija 2004
Computer Arithmetic
19
Carry-Skip Adder:
N-bits, k-bits/group, r=N/k groups
...
O R
C
out
a
N - 1 b N - 1 N - k - 1b N - k - 1
N -1
b (r-1 )k a
O R
S
P
...
G
2 k -1
O R
...
r1
N -k -1
r-1
AND
P
AND
2 k -1
ak
bk
k -1
r-2
(r-2 )k
2 k -1
k -1
O R
S
P
...
a0 b
...
G
... ...
... ...
S
(r-1 )k -1
...
... ...
... ...
S
(r-1 )k
k -1
S
P
AND
0
0
AN D
c r it ic a l p a t h , d e la y = 2 ( k - 1 ) + ( N / 2 - 2 )
Oklobdzija 2004
Computer Arithmetic
20
in
Carry-Skip Adder
tp
ripple adder
bypass adder
t d 2 k 1 t RCA
2 t SKIP
2k
4..8
Oklobdzija 2004
N
Computer Arithmetic
21
Oklobdzija 2004
Computer Arithmetic
22
N -1
...
..
C
out
ajb
..
.
N -1
N -1
m -1
S
P
m -1
m -1
a0 b
...
G
m -2
m -2
m -2
...
...
..
in
s k ip in g
...
C
ou
r ip p lin g
Oklobdzija 2004
C a r r y s ig n a l p a th
Computer Arithmetic
23
in
1 3
6
5
=9
3 1
Any-point-to-any-point delay = 9
as compared to 12 for CSKA
Oklobdzija 2004
Computer Arithmetic
24
Oklobdzija 2004
Computer Arithmetic
25
G0
P1
P2
G1
P3
G2
BP
Co,3
G3
BP
Delay model:
Oklobdzija 2004
Computer Arithmetic
26
t d c1 c2 N c3
Oklobdzija, Barnes, Arith85
Oklobdzija 2004
Computer Arithmetic
27
Oklobdzija 2004
Computer Arithmetic
28
Oklobdzija 2004
Computer Arithmetic
29
16
VBA
14
12
CLA
10
8
VBA- Multi-Level
6
4
2
0
4
11
18
25
32
39
46
53
60
Size N
Oklobdzija 2004
Computer Arithmetic
30
VLSI Arithmetic
Lecture 4
Prof. Vojin G. Oklobdzija
University of California
http://www.ece.ucdavis.edu/acsel
Review
Lecture 3
Oklobdzija 2004
Computer Arithmetic
33
..
...
Gm
Gm-1
SN-1
Pm
Pm-1
Gm
Gm-1
aj b j
..
.
Gm-2
a 0 b0
...
...
..
G2
G1
G0
Sj
Si
Pm-2
P2
Gm-2
bi
...
Cin
S0
P1
G2
G1
skiping
P0
G0
...
C out
Cin
rippling
Oklobdzija 2004
Computer Arithmetic
34
1 3
6
5
=9
3 1
Any-point-to-any-point delay = 9
as compared to 12 for CSKA
Oklobdzija 2004
Computer Arithmetic
35
Oklobdzija 2004
Computer Arithmetic
36
G0
P1
P2
G1
P3
G2
BP
Co,3
G3
BP
Delay model:
Oklobdzija 2004
Computer Arithmetic
37
t d c1 c2 N c3
Oklobdzija, Barnes, Arith85
Oklobdzija 2004
Computer Arithmetic
38
Oklobdzija 2004
Computer Arithmetic
39
Oklobdzija 2004
Computer Arithmetic
40
16
Square Root
Dependency
14
VBA
12
Log
Dependency
10
CLA
VBA- Multi-Level
6
4
2
0
4
11
18
25
32
39
46
53
60
Size N
Oklobdzija 2004
Computer Arithmetic
41
Circuit Issues
Adder speed can not be estimated based
on:
logic gates in the critical path
number of transistors in the path
logic levels in the path
Computer Arithmetic
42
Fan-Out Dependency
Oklobdzija 2004
Computer Arithmetic
43
Fan-In Dependency
This looks like
Logical Effort
(1985)
Oklobdzija 2004
Computer Arithmetic
44
Oklobdzija 2004
Computer Arithmetic
45
Oklobdzija 2004
Computer Arithmetic
46
Carry-Lookahead Adder
(Weinberger and Smith, 1958)
Computer Arithmetic
47
g i ai bi
pi ai bi
c out
s
c in
si pi ci
si
Oklobdzija 2004
Computer Arithmetic
48
CLA
Definitions:
4-bit
Adder
a
a
a
b
b
b
b
ai+3
i+2
i+3
Ci+4
Ci+3
gi+3
pi+3
i+1
i+2
Ci+2
gi+2
pi+2
i+1
Ci+1
gi+1
pi+1
Ci
gi
pi
ci 1 ai bi ci ai bi ci ai bi g i pi ci
ci 2 g i 1 pi 1ci 1 g i 1 pi 1 ( g i pi c1 )
g i 1 pi 1 g i pi 1 pi c1
Oklobdzija 2004
Computer Arithmetic
49
Carry-Lookahead
a
a 4-bits
a
a Adder:
b
b
b
b
i+3
i+2
i+3
Ci+4
Ci+3
gi+3
pi+3
i+1
i+2
Ci+2
gi+2
pi+2
i+1
Ci+1
gi+1
pi+1
Ci
gi
pi
ci 3 g i 2 pi 2 ci 2 g i 2 pi 2 ( g i 1 pi 1 g i pi 1 pi ci )
g i 2 pi 2 g i 1 pi 2 pi 1 g i pi 2 pi 1 pi ci
ci 4 g i 3 pi 3ci 3 g i 3 pi 3 ( g i 2 pi 2 g i 1 pi 2 pi 1 g i )
g i 3 pi 3 g i 2 pi 3 pi 2 g i 1 pi 3 pi 2 pi 1 g i pi 3 pi 2 pi 1 pi ci
Gj
Oklobdzija 2004
Computer Arithmetic
Pj
50
Carry-Lookahead Adder
G j g i 3 pi 3 g i 2 pi 3 pi 2 g i 1 pi 3 pi 2 pi 1 g i
Pj pi 3 pi 2 pi 1 pi
i+ 3
i+ 2
i+ 2
i+ 1
i+ 1
C in
g
C
i+ 1p i+ 1
4 (j+ 1 )
i+ 1 p i+ 1
i+ 1
i+ 1
P , G G ro u p
c4( j 1) G j Pj c j
Oklobdzija 2004
i+ 3
4 j+ 3
4 j+ 2
4 j+ 1
Computer Arithmetic
51
Carry-Lookahead Adder
(Weinberger and Smith)
P * j Pi 3 Pi 2 Pi 1 Pi
G
j+ 3
j+ 3
j+ 2
j+ 2
j+ 1
j+ 1
4 (j+ 1 )
P*
G*
c4 ( j 1) G *k P *k c4 j
4 j+ 3
4 j+ 2
4 j+ 1
Computer Arithmetic
52
4j
28
24
in d iv id u a l a d d e r s
g e n e r a t i n g : g i, p i,
and sum S i
20
12
C a r r y - lo o k a h e a d s u p e r - b lo c k s o f
4 - b it s b lo c k s g e n e r a tin g :
G * i, P * i, a n d C in f o r t h e 4 - b i t
b lo c k s
out
16
out
in
in
in
C a r r y - lo o k a h e a d b lo c k s o f
4 - b it s g e n e r a t in g :
G i, P i, a n d C in f o r t h e
a d d e rs
G r o u p p r o d u c in g f in a l
c a rry C out a n d C 16
C r i t i c a l p a t h d e l a y = ( f o r g i , p i ) + 2 x 2 ( f o r G , P ) + 3 x 2 ( f o r C i n ) + 1 X O R - ( f o r S u m ) = a p p x . 1 2 o f d e l a y
Oklobdzija 2004
Computer Arithmetic
53
Carry-Lookahead Adder
(Weinberger and Smith: original derivation, 1958 )
Oklobdzija 2004
Computer Arithmetic
54
Carry-Lookahead Adder
(Weinberger and Smith: original derivation )
Oklobdzija 2004
Computer Arithmetic
55
Oklobdzija 2004
Computer Arithmetic
56
Oklobdzija 2004
Computer Arithmetic
57
PG BLO C K
PG BLO CK
P63:60
P63:48
P63:0
P59:48
G59:48
P55:48
G55:48
P51:48
C4
C8
C12
C16
C32
C48
C52
C56
G63:0
C60
P47:32
G63:48
61
P59:56
G59:56
P55:52
G55:52
P51:48
G51:48
P11:0
G11:0
P7:0
G7:0
P3:0
G3:0
G15:0
P15:0
G31:16
P31:16
G47:32
G63:60
C
P,G62:60 63
C
P,G61:6062
C
P,G60
CAR RY
BLO C K
P63
G63
P62
G62
P61
G61
P60
G60
P59
PG BLO CK
PG BLO C K
G56
P55
G52
G51:48
P15:12
G15:12
P11:8
G11:8
P7:4
G7:4
P3:0
G3:0
C16
P15:0
G31:0
C32
P31:0
G47:0
P47:0
C48
3.75nS
G15:0
C64
63
- S
63
- C
60
P51
G48
P,G2:0
P,G1:0
P,G0
C0
2.35nS
2.0nS
P47
G32
P
31
G16
P15
PG BLO CK
1.7nS
- C
48
- C
4 7 :0
- G
1 5 :0
- G
3 :0
- G
0
C r itic a l p a th : A , B - G
59
Computer Arithmetic
Oklobdzija 2004
G12
P11
G8
P7
G4
P3
G3
P2
G2
P1
G1
P0
G0
1.05nS
...
...
...
...
...
...
...
...
4.8nS
2.7nS
Motorola's 64-bit
CLA
conventional PG Block
no better
situation here !
Computer Arithmetic
60
Motorola's 64-bit
CLA
Modified PG Block
Oklobdzija 2004
Computer Arithmetic
61
3.9nS
1.8nS
2.2nS
3.55nS
2.9nS
Oklobdzija 2004
3.2nS
Computer Arithmetic
62
P6 3
G6 3
P6 2
G6 2
P6 1
G6 1
P6 0
G
60
P5 9
G
56
P5 5
G5 2
P5 1
G4 8
P4 7
G3 2
P3 1
G1 6
P
15
G1 2
P1 1
G8
P7
P6 3 : 6 0
P6 3 : 4 8
P6 3 : 0
G5 5 : 4 8
P5 5 : 4 8
C5 6
G5 9 : 4 8
P5 9 : 4 8
C6 0
G6 3 : 0
P4 7 : 3 2
G6 3 : 4 8
C6 1
P5 9 : 5 6
G5 9 : 5 6
P5 5 : 5 2
G5 5 : 5 2
P5 1 : 4 8
C
16
C3 2
C4 8
C5 2
P5 1 : 4 8
G5 1 : 4 8
G3 : 0
P3 : 0
G7 : 0
P7 : 0
C4
C8
G1 1 : 0
P1 1 : 0
C1 2
G5 1 : 4 8
P1 5 : 1 2
G1 5 : 1 2
P1 1 : 8
G1 1 : 8
P7 : 4
G7 : 4
P3 : 0
G3 : 0
G1 5 : 0
P1 5 : 0
G3 1 : 1 6
P3 1 : 1 6
G4 7 : 3 2
C G6 3 : 6 0
P , G6 2 : 6 0 6 3
C
P , G6 1 : 66 02
P , G6 0
P , G2 : 0
P , G1 : 0
P , G0
C4 8
P4 7 : 0
G4 7 : 0
C3 2
P3 1 : 0
G3 1 : 0
C1 6
P1 5 : 0
G1 5 : 0
C6 4
C0
63
3.2nS
2.9nS Computer Arithmetic
Oklobdzija 2004
3.55nS
2.2nS
3.9nS
2.35nS
2.0nS
6 3
- S
6 3
- C
6 0
- C
4 8
- C
4 7 :0
- G
1 5 :0
- G
3 :0
- G
0
- G
C r it ic a l p a t h : A , B
B L O C K
1.7nS
P G
3.75nS
2.7nS
B L O C K
P G
B L O C K
P G
1.05nS
G4
P3
G3
P2
G2
P1
G1
P0
G0
C A R R Y
B L O C K
B L O C K
P G
B L O C K
P G
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
4.8nS
1.8nS
Delay
Optimized
CLA: LeeOklobdzija 91
(a.) Fixed groups and levels
(b.) variable-sized groups,
fixed levels
(c.) variable-sized groups and
fixed levels
(d.) variable-sized groups and
levels
Oklobdzija 2004
Computer Arithmetic
65
Oklobdzija 2004
Computer Arithmetic
66
Oklobdzija 2004
Computer Arithmetic
67
Oklobdzija 2004
Computer Arithmetic
68
Oklobdzija 2004
Computer Arithmetic
69
70
Computer Arithmetic
71
Lings Adder
Huey Ling, High-Speed Binary Adder
IBM Journal of Research and Development, Vol.5, No.3, 1981.
Used in: IBM 3033, IBM 168, Amdahl V6, HP etc.
Lings Derivations
ai
bi
define:
Ci 1 g i pi Ci
H i 1 Ci 1 Ci
ci+1
ci
g i ai bi
si
pi Ci pi Ci 1 pi g i pi pi Ci 1
ai bi pi gi
ti
pi Ci 1 pi Ci 1 pi H i 1
0 0 0
1 1 0
0 1 0
1 0 1
pi Ci pi H i 1
Ci 1 ti H i 1
Ci 1 gi pi Ci gi H i 1 pi Ci
gi H i 1 pi H i 1 ti H i 1
Oklobdzija 2004
Computer Arithmetic
73
From:
Lings Derivations
and
CiC1
gi
pi CCi
H ii 11 Cii 11 Cii g i p
g
i i
i
i
i
because:
H i 1 g i ti 1 H i
Ci 1 ti H i 1
fundamental expansion
Oklobdzija 2004
Computer Arithmetic
74
Ling Adder
Variation of CLA:
Lings equations:
pi ai bi
ti ai bi
gi ai bi
gi ai bi
Ci 1 g i pi Ci
H i 1 g i ti 1 H i
Si pi Ci
Si ti H i 1 g i ti 1 H i
Ling, IBM J. Res. Dev, 5/81
Oklobdzija 2004
Computer Arithmetic
75
Ling Adder
Lings equation:
Variation of CLA:
Ci 1 g i g i Ci pi Ci
g i g i pi Ci
Ci 1 g i ti Ci
H i g i ti 1 H i 1
Ling uses different transfer function.
Four of those functions have desired
properties (Lings is one of them)
see: Doran, IEEE Trans on Comp. Vol 37, No.9 Sept. 1988.
Oklobdzija 2004
Computer Arithmetic
76
Ling Adder
Conventional:
Fan-in of 5
Oklobdzija 2004
Computer Arithmetic
77
Computer Arithmetic
78
VLSI Arithmetic
Lecture 5
Prof. Vojin G. Oklobdzija
University of California
http://www.ece.ucdavis.edu/acsel
Review
Lecture 4
Lings Adder
Huey Ling, High-Speed Binary Adder
IBM Journal of Research and Development, Vol.5, No.3, 1981.
Used in: IBM 3033, IBM S370/168, Amdahl V6, HP etc.
Lings Derivations
ai
bi
define:
Ci 1 g i pi Ci
H i 1 Ci 1 Ci
ci+1
ci
g i ai bi
si
pi Ci pi Ci pi g i pi pi Ci
ai bi pi gi
ti
pi Ci pi Ci 1 pi H i 1
0 0 0
1 1 0
0 1 0
1 0 1
pi Ci pi H i 1
Ci 1 ti H i 1
Ci 1 gi pi Ci gi H i 1 pi Ci
gi H i 1 pi H i 1 ti H i 1
Oklobdzija 2004
Computer Arithmetic
82
From:
Lings Derivations
and
CiC1
gi
pi CCi
H ii 11 Cii 11 Cii g i p
g
i i
i
i
i
because:
H i 1 g i ti 1 H i
Ci 1 ti H i 1
fundamental expansion
Oklobdzija 2004
Computer Arithmetic
83
Ling Adder
Variation of CLA:
Lings equations:
pi ai bi
ti ai bi
gi ai bi
gi ai bi
Ci 1 g i pi Ci
H i 1 g i ti 1 H i
Si pi Ci
Si ti H i 1 g i ti 1 H i
Ling, IBM J. Res. Dev, 5/81
Oklobdzija 2004
Computer Arithmetic
84
Ling Adder
Lings equation:
Variation of CLA:
Ci 1 g i g i Ci pi Ci
g i g i pi Ci
ai
bi
Hi+1
ci+1
ai-1 bi-1
Hi
g i, t i
ci
si
Ci 1 g i ti Ci
gi-1, ti-1
ci-1
si-1
H i 1 g i ti 1 H i
Ling uses different transfer function.
Four of those functions have desired
properties (Lings is one of them)
see: Doran, IEEE Trans on Comp. Vol 37, No.9 Sept. 1988.
Oklobdzija 2004
Computer Arithmetic
85
Ling Adder
Conventional:
Fan-in of 5
Oklobdzija 2004
Computer Arithmetic
86
Advantages
of and
Lings
Uniform
loading in fan-in
fan-outAdder
H16 contains 8 terms as compared to G16 that
contains 15.
H16 can be implemented with one level of logic
(in ECL), while G16 can not (with 8-way wire-OR).
(Lings adder takes full advantage of wired-OR, of
special importance when ECL technology is
used - his IBM limitation was fan-in of 4 and
wire-OR of 8)
Oklobdzija 2004
Computer Arithmetic
87
Oklobdzija 2004
Computer Arithmetic
88
Oklobdzija 2004
Computer Arithmetic
89
Oklobdzija 2004
Computer Arithmetic
90
Computer Arithmetic
91
Implementation of Lings
Adder in CMOS
(S. Naffziger, A Subnanosecond 64-b Adder, ISSCC 96)
Oklobdzija 2004
Computer Arithmetic
92
S. Naffziger,
ISSCC96
H 4 g 3 g 2 t 2 g1 t 2t1 g 0
Ci 1 ti H i 1
Oklobdzija 2004
Computer Arithmetic
93
S. Naffziger,
ISSCC96
H 4 g 3 g 2 t 2 g1 t 2t1 g 0
Oklobdzija 2004
Computer Arithmetic
94
H 4 g 3 g 2 t 2 g1 t 2t1 g 0
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
95
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
96
S. Naffziger, ISSCC96
Oklobdzija 2004
Computer Arithmetic
97
S. Naffziger, ISSCC96
Oklobdzija 2004
Computer Arithmetic
98
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
99
Computer Arithmetic
100
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
101
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
102
S. Naffziger,
ISSCC96
Oklobdzija 2004
Computer Arithmetic
103
Oklobdzija 2004
Computer Arithmetic
104
CK
A2
CK
A3
B2
A2
A1
B2
B1
A1
G4
B0
G0
P1
G1
B1 A3
B3
A0
B0 A2
B2
P4
CK
LC
CK
A1
B3
B1
A0
CK
G2
P2
CK
SumL
K
G
C1L
LCH
LCL
C1H C0L
C0H
P
C1H
SumH
C1L C0H
LCH
C0L
LCL
CK
Oklobdzija 2004
Computer Arithmetic
105
LCS4 Critical G
Path
in1
4b
(k,p) or (g,p)
P4
G3
G4
12b
C15
32b
C47
C31
C15
16b
S63 S62
Oklobdzija 2004
S48
Computer Arithmetic
106
Stages
dg3# (dg3)
g4 (NAND2)
C15# (GG4)
C15 (INV)
C47# (LC)
C47 (INV)
C47#b (INV)
C47b (INV)
S63# (SUM)
S63 (INV)
Branch
4.0
2.0
1.0
1.0
3.0
1.0
1.0
1.0
16.0
1.0
Oklobdzija 2004
LE
0.98
1.11
1.01
1.00
1.03
1.00
1.00
1.00
0.86
1.00
Total
Path
Effort
fo, opt
Parasitic Branch Total LE
2.97
1.84
1.80
1.00
3.32
3.84E+02 9.73E-01 3.74E+02 1.81
1.00
1.00
1.00
1.36
1.00
Computer Arithmetic
Effort
Delay
(ps)
Parasitic
Delay
(ps)
Total
Delay
(ps)
Total
Delay
(FO4)
66
70
136
7.2
107
Results:
0.5u Technology
Speed: 0.930 nS
Nominal process, 80C, V=3.3V
Oklobdzija 2004
Computer Arithmetic
108
Prefix Adders
and
Parallel Prefix Adders
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
110
Prefix Adders
(g0, p0)
G i, P i =
i=0
ci+1 = Gi
c1 = g0+ p0 cin
1in
for i=0, 1, .. n
(g-1, p-1)=(cin,cin)
Computer Arithmetic
111
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
112
Oklobdzija 2004
Computer Arithmetic
113
Pyramid Adder:
M. Lehman, A Comparative Study of Propagation Speed-up Circuits in Binary Arithmetic
Units, IFIP Congress, Munich, Germany, 1962.
Oklobdzija 2004
Computer Arithmetic
114
Oklobdzija 2004
Computer Arithmetic
115
Oklobdzija 2004
Computer Arithmetic
116
Oklobdzija 2004
Computer Arithmetic
117
Oklobdzija 2004
Computer Arithmetic
118
Computer Arithmetic
119
Computer Arithmetic
120
Oklobdzija 2004
Computer Arithmetic
121
Kogge-Stone Adder
Oklobdzija 2004
Computer Arithmetic
122
Computer Arithmetic
123
Brent-Kung Adder
Oklobdzija 2004
Computer Arithmetic
124
Oklobdzija 2004
Computer Arithmetic
125
Oklobdzija 2004
Computer Arithmetic
126
Computer Arithmetic
127
Computer Arithmetic
128
Knowles 1999
Computer Arithmetic
129
Computer Arithmetic
130
Computer Arithmetic
131
Conclusion
Irregular, hybrid schmes are
possible
The speed-up of 15% is
achieved at the cost of large
wiring, hence area and power
Circuits close in speed to K-S
are available at significantly
lower wiring cost
Oklobdzija 2004
Computer Arithmetic
132
VLSI Arithmetic
Lecture 6
Prof. Vojin G. Oklobdzija
University of California
http://www.ece.ucdavis.edu/acsel
Review
Lecture 5
Prefix Adders
and
Parallel Prefix Adders
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
136
Prefix Adders
(g0, p0)
G i, P i =
i=0
ci+1 = Gi
c1 = g0+ p0 cin
1in
for i=0, 1, .. n
(g-1, p-1)=(cin,cin)
Computer Arithmetic
137
Oklobdzija 2004
Computer Arithmetic
138
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
139
Oklobdzija 2004
Computer Arithmetic
140
Oklobdzija 2004
Computer Arithmetic
141
Oklobdzija 2004
Computer Arithmetic
142
Kogge-Stone Adder
Oklobdzija 2004
Computer Arithmetic
143
Brent-Kung Adder
Oklobdzija 2004
Computer Arithmetic
144
Oklobdzija 2004
Computer Arithmetic
145
Pyramid Adder:
M. Lehman, A Comparative Study of Propagation Speed-up Circuits in Binary Arithmetic
Units, IFIP Congress, Munich, Germany, 1962.
Oklobdzija 2004
Computer Arithmetic
146
Computer Arithmetic
147
Computer Arithmetic
148
Oklobdzija 2004
Computer Arithmetic
149
Computer Arithmetic
150
Han-Carlson
151
Oklobdzija 2004
Computer Arithmetic
152
Oklobdzija 2004
Computer Arithmetic
153
Computer Arithmetic
154
Computer Arithmetic
155
Knowles 1999
Computer Arithmetic
156
Computer Arithmetic
157
Computer Arithmetic
158
Conclusion
Irregular, hybrid schmes are
possible
The speed-up of 15% is
achieved at the cost of large
wiring, hence area and power
Circuits close in speed to K-S
are available at significantly
lower wiring cost
Oklobdzija 2004
Computer Arithmetic
159
Computer Arithmetic
160
Oklobdzija 2004
Computer Arithmetic
161
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
163
Conditional
Sum Adder
Oklobdzija 2004
Computer Arithmetic
164
from: Ercegovac-Lang
Oklobdzija 2004
Computer Arithmetic
165
Oklobdzija 2004
Oklobdzija 2004
Computer Arithmetic
167
Carry-Select Adder
O. J. Bedrij, Carry-Select Adder, IRE
Transactions on Electronic Computers, June
1962, p.340-34
Oklobdzija 2004
Carry-Select Adder
Addition under assumption of C in=0 and Cin =1.
Oklobdzija 2004
Computer Arithmetic
170
Computer Arithmetic
171
Carry-Select Adder
Computer Arithmetic
172