LOGIC OPTIMIZATION USING TECHNOLOGY INDEPENDENT MUX BASED ADDERS IN FPGA

International Journal of VLSI design & Communication Systems (VLSICS) Vol.3, No.4, August 2012
DOI : 10.5121/vlsic.2012.3412 135
LOGIC OPTIMIZATION USING TECHNOLOGY
INDEPENDENT MUX BASED ADDERS IN FPGA
R.Uma and P.Dhavachelvan
Department of Computer Engineering, School of Engineering, Pondicherry University,
Pondicherry, India
uma.ramadass1@gmail.com
dhavachelvan@gmail.com
ABSTRACT
Adders form an almost obligatory component of every contemporary integrated circuit. The prerequisite of
the adder is that it is primarily fast and secondarily efficient in terms of power consumption and chip area.
Therefore, careful optimization of the adder is of the greatest importance. This optimization can be attained
in two levels; it can be circuit or logic optimization. In circuit optimization the size of transistors are
manipulated, where as in logic optimization the Boolean equations are rearranged (or manipulated) to
optimize speed, area and power consumption. This paper focuses the optimization of adder through
technology independent mapping. The work presents 20 different logical construction of 1-bit adder cell in
CMOS logic and its performance is analyzed in terms of transistor count, delay and power dissipation.
These performance issues are analyzed through Tanner EDA with TSMC MOSIS 250nm technology. From
this analysis the optimized equation is chosen to construct a full adder circuit in terms of multiplexer. This
logic optimized multiplexer based adders are incorporated in selected existing adders like ripple carry
adder, carry look-ahead adder, carry skip adder, carry select adder, carry increment adder and carry save
adder and its performance is analyzed in terms of area (slices used) and maximum combinational path
delay as a function of size. The target FPGA device chosen for the implementation of these adders was
Xilinx ISE 12.1 Spartan3E XC3S500-5FG320. Each adder type was implemented with bit sizes of: 8, 16,
32, 64 bits. This variety of sizes will provide with more insight about the performance of each adder in
terms of area and delay as a function of size.
KEYWORDS
Technology independent mapping, Adder topologies, FPGA, Multiplexer based Adders, Logical effort,
Delay calculation
1. INTRODUCTION
In most of the digital systems, adders are the fundamental component in the design of application
specific integrated circuits like RISC processors, digital signal processors (DSP), microprocessors
etc. The design criterion of a full adder cell is usually multi-fold. Transistor count is, of course, a
primary concern which largely affects the design complexity of many function units such as
multiplier and Arithmetic logic unit (ALU). The basic principle in designing digital adder
circuit hovers around reducing the required hardware thus reducing the cost too. To achieve
this, logical optimization helps to obtaining minimum number of literals to minimizing the
transistor count and the power consumption and increasing the speed of operation.
A logic expression can be expressed in various logic forms, which differ in literal counts.
In widely-used MOS circuits, the number of transistors to implement a Boolean expression is
directly proportional to literal counts in its logic form[1-2]. Thus, a logic optimization is
simply to derive a logic form with the fewest literals. Logic level optimization is the design
task where an RTL circuit description is optimized in terms of area, delay and power.

136
Conventionally, a logic level optimization can be achieved in two steps; they are Technology
Independent (TI) and Technology – Dependent (TD) optimization. In the former method the
circuit’s Boolean description is optimized ignoring the technology in which the circuit will be
implemented. In the second method, the output of the technology independent optimization step
(i.e. optimized Boolean network) is optimized considering the adopted technology. During the TI
step there is much flexibility to restructure circuit logic to minimize the number of nodes and
literals, thereby reducing the area of the circuit. During this stage the circuit can be most
effectively restructured to meet the specified delay constraints critical for circuit performance.
During the TD step, the delay characteristics of the target library are available, but very few
restructuring of the circuit is possible.
Logical effort [13, 14] has been widely used in a variety of application domains as well as in
industry standard EDA synthesis tools. Designing a circuit to achieve the greatest speed or to
meet a delay constraint presents a bewildering array of choices [13, 14]. The method of logical
effort is a design procedure for achieving the least delay along a path of a logic network. This
method is based on a simple approximation that treats MOS circuits as networks of resistance and
capacitance. This RC model provides simple mathematical calculation to obtain the circuit’s
maximum speed. In this paper the delay model for optimized full adder circuit and its delay
estimation is also presented.
In this paper, we proposed 20 different Boolean expressions (logic construction) to implement a
1-bit full adder circuit. All the Boolean expressions are realized in terms of CMOS logic. The
optimization method used in this work is technology independent optimization step. These
Boolean logic realization and performances are analyzed in terms of transistor count, delay and
power dissipation using Tanner EDA with TSMC MOSIS 250nm technology. From this analysis
the optimized equation is selected and it is implemented in terms of multiplexers and it is
incorporated in selected existing adder topologies like ripple carry adder, carry look-ahead adder,
carry skip adder, carry select adder, carry increment adder and carry save adder and its
performance is analyzed in terms of area (slices used) and maximum combinational path delay as
a function of size. Performance comparison of existing and logic optimized schemes are analyzed
on cell-based VLSI technologies, such as standard-cell based FPGAs. The cell-based approach is
justiﬁed by its wide-spread use in the ASIC design community and its compatibility with
hardware synthesis, which in turns satisﬁes the demand for ever higher productivity. This work
presents the significance of adder comparison in terms of CLBs occupied and its maximum
combinational delay exist in adder topology.
The organization of the paper is as follows: The section 2, describes the existing adder topologies.
Section 3, presents the mathematical Boolean expression for the design of 1-bit full adder cell.
Section 4 presents the simulation and analysis of full adder using Tanner EDA. Section 5 presents
the FPGA implementation of different adder topologies. Section 6 gives the summary of
comparison. Finally the conclusion is presented in section 7.
2. REVIEW OF EXISTING ADDER TOPOLOGY
Most of the VLSI applications, such as digital signal processing, image and video processing, and
microprocessors, extensively use arithmetic operations. Addition, subtraction, multiplication, and
multiply and accumulate (MAC) are examples of the most commonly used operations. The 1-bit
full-adder cell is the building block of all these modules. Thus, enhancing its performance is
critical for enhancing the overall module performance. This section presents the overview of the
existing adder topologies.
In FPGAs, the most fundamental component implemented for high speed applications like
microprocessors, arithmetic logic unit, program counters and multiply accumulate unit. Lot of

137
implementations has been made for these adder topologies for optimizing area, delay and power
dissipations. In the reference [1], it provides an overview for the comparison of adders in the
early design phase for selecting their appropriate design structure for implementing adders with
the constraints of area, delay and power dissipation. This paper also reveals the pre-estimation of
energy-delay, product, energy-delay estimation and power estimation in the energy delay space.
In the reference [2], the proposed high speed and low power full adder cells which has designed
with pass transistor logic styles to reduce the power delay product (PDP). This paper also reports
the performance comparison of adder cells with CMOS, DCVS, CPL, DPL, Swing restorer CPL
and hybrid styles. This paper shows the implementation of adder cells with enhanced carry
generation stage which is implemented with multiplexes. This feature provides that for this logic
there are no internal signals being generated for controlling the selection of output multiplexers,
thereby reducing the full voltage swing, delay and overall propagation delays.
The adder topology is present in literature [3-12], Ripple Carry Adder (RCA) is the simplest, but
slowest adders with O(n) area and O(n) delay, where n is the operand size in bits. Carry Look-
Ahead (CLA) have O(nlog(n)) area and O(log(n)) delay, but typically suffer from irregular
layout. On the other hand, Carry Skip Adder, carry increment and carry select have O(n) area and
2/ 1
( )l l
O n + +
delay provides a good compromise in terms of area and delay, along with a simple
and regular layout. Carry save adder have O(n) area and O(log n) delay. The ripple carry adder,
the most basic of flavours, is at the one extreme of the spectrum with the least amount of CLBs
but the highest delay. CLA adders can be realized in two gate levels provided there is no limit on
fan in/out. The carry select adders reduce the computation time by pre-computing the sum for all
possible carry bit values (ie ‘0’ and ‘1’). After the carry becomes available the correct sum is
selected using multiplexer. Carry select adders are in the class of fast adders, but they suffer from
fan-out limitation since the number of multiplexers that need to be driven by the carry signal
increases exponentially. In the worst case, a carry signal is used to select n/2 multiplexers in an n-
bit adder. When three or more operands are to be added simultaneously using two operand adders,
the time consuming carry propagation must be repeated several times. If the number of operands
is ‘k’, then carries have to propagate (k-1) times.
3. MATHEMATICAL EQUATIONS FOR FULL ADDER
A full adder is a combinational circuit that performs the arithmetic sum of three bits: A, B and a
carry in, C, from a previous addition produces the corresponding SUM, S, and a carry out,
CARRY. The various equations for SUM and CARRY are given below
ܷܵ‫ܯ‬ = ‫ܣ‬ ⊕ ‫ܤ‬ ⊕ ‫ܥ‬ (1)
ܷܵ‫ܯ‬ = ቀ‫ܤܣ‬ + ‫ܣܤ‬ቁ ‫ܥ‬ + ൫‫ܤܣ‬ + ‫ܤܣ‬൯‫ܥ‬ (3)
‫ܻܴܴܣܥ‬ = (‫ܣ‬ ⊕ ‫ܥ)ܤ‬ + ‫ܤܣ‬ (5)
‫ܻܴܴܣܥ‬ = ൫‫ܤܣ‬ + (‫ܣ‬ ⊕ ‫ܤ‬൯‫ܥ‬ (7)
‫ܻܴܴܣܥ‬ = ‫ܣ‬ ⊕ ‫ܤ‬ • ‫ܤ‬ + ‫ܣ‬ ⊕ ‫ܤ‬ • ‫ܥ‬ (9)
‫ܻܴܴܣܥ‬ = ‫ܥ‬ • (‫ܣ‬ • ‫)ܤ‬ + ‫ܥ‬ • (‫ܣ‬ + ‫)ܤ‬ (11)
‫ܻܴܴܣܥ‬ = (‫ܣ‬ ⊕ ‫ܥ)ܤ‬ • ‫ܤܣ‬ (13)
‫ܻܴܴܣܥ‬ = ‫ܤܣ‬ • ‫ܥܣ‬ • ‫ܥܤ‬ (15)
‫ܻܴܴܣܥ‬ = ‫ܤܣ‬ • ‫ܣ‬ + ‫ܤ‬ • ‫ܥ‬ (17)
‫ܻܴܴܣܥ‬ = ‫ܣ‬ + ‫ܤ‬ + ‫ܣ‬ ⊕ ‫ܤ‬ + ‫ܥ‬ (19)
‫ܻܴܴܣܥ‬ = ൫‫ܣ‬ ⊕ ‫ܤ‬൯‫ܥ‬ ⊕ ‫ܤܣ‬ (20)
ܷܵ‫ܯ‬ = ‫ܣ‬ ⊕ ‫ܤ‬ ⊕ ‫ܥ‬ (2)
ܷܵ‫ܯ‬ = ቀ‫ܣ‬ ‫ܤ‬ + ‫ܤܣ‬ቁ ‫ܥ‬ + ൫‫ܣ‬ ‫ܤ‬ + ‫ܤܣ‬൯‫ܥ‬ (4)
‫ܻܴܴܣܥ‬ = ൬‫ܣ‬ ⊕ ‫ܤ‬൰ ‫ܥ‬ + ‫ܤܣ‬ (6)
‫ܻܴܴܣܥ‬ = ‫ܤܣ‬ • ‫ܥܣ‬ • ‫ܥܤ‬ (8)
‫ܻܴܴܣܥ‬ = (‫ܣ‬ ⊕ ‫)ܤ‬ • ‫ܥ‬ ⊕ ‫ܤܣ‬ (10)
‫ܻܴܴܣܥ‬ = ‫ܤܣ‬ + ‫ܥܣ‬ + ‫ܥܤ‬ (12)
‫ܻܴܴܣܥ‬ = ‫ܤܣ‬ + ‫ܣ‬ ⊕ ‫ܤ‬ • ‫ܥ‬ (14)
‫ܻܴܴܣܥ‬ = ‫ܣ‬ ⊕ ‫ܤ‬ • ‫ܥܣ‬ • ‫ܥܤ‬ (16)
‫ܻܴܴܣܥ‬ = ‫ܣ‬ ⊕ ‫ܤ‬ • ‫ܥ‬ ⊕ ‫ܤܣ‬ (18)

138
Figure 1 Different structure for realizing Full Adder circuit
In this work 20 different Boolean expressions are formulated (Figure 1 shows different full adder
circuit). Using this logical equation it is possible to construct 64 full adder circuits. These adders

139
are implemented with CMOS logic with technology independent optimization process and its
performance are analyzed in terms of transistor count, delay and power dissipation using Tanner
EDA with TSMC MOSIS 250nm technology. From this analysis the optimized equation is
selected and it is implemented in terms of multiplexers and it is incorporated in selected existing
adder topologies like ripple carry adder, carry look-ahead adder, carry skip adder, carry select
adder, carry increment adder and carry save adder and its performance is analyzed in terms of
area (slices used) and maximum combinational path delay as a function of size.
Mathematically it is also possible to calculate the delay of a circuit by constructing delay models
instead of simulation tools using logical effort methods. The logical effort provides a simple
method “on the back of an envelope” [13, 14] to choose the best topology (logical constructs) and
number of stages of logic for a function. Speed optimization of circuit network can be achieved
by the method of logical effort. This method provides how many stages of logic are required for
the fastest implementation of any given logic function. The speed of the circuit depends on the
capacitive load that the circuit of the logic gate drives and the logic function of the gate. The
delay incurred in a logic gate are expressed as sum of two components namely, the parasitic delay
p and the effort delay f as follows [13-14].
The delay in single stage network is expressed as
d = gh + p (1)
where,
g - Logical effort (the ability of the logic gate’s topology to produce output current)
h - Electrical effort (the ratio of output capacitance to input capacitance)
p - Intrinsic delay (delay of the gate due to its own internal capacitance)
Table 1 presents the logical effort of common static CMOS gates assuming the aspect ratio of
pull-up and pull-down network to be 2:1 to have equal rise and fall delay. Table 2 presents the
parasitic delay of CMOS logic independent of the size of the logic gate and of the load
capacitance it drives. The principle contribution to parasitic delay is the capacitance of the
source/drain regions of the transistors that drive the gate’s output.
An example to calculate the delay of a full adder is shown in Figure (2) using the expression
SU M A B C= ⊕ ⊕ and C A R R Y A B A C B C= • • . The circuit is realized as two stage
network, stage1 and stage2 respectively. Assume that the input capacitance of 10pf on each input
and it will drive the output capacitance with a maximum of 10pf.
Table 1. Logical Effort of CMOS gates Table 2. Parasitic delay of CMOS gates
Gate type
Number of Inputs
Gate type
Number of Inputs
1 2 3 4 5 n 1 2 3 4 5 n
Inverter 1 Inverter 1
NAND 4/3 5/3 6/3 7/3 (n+1)/3 NAND 2 3 4 5 nPinv
NOR 5/3 7/3 9/3 11/3 (2n+1)/3 NOR 2 3 4 5 nPinv
XOR/XNOR 4 12 32 XOR/XNOR 4 4 4 4Pinv

140
Delay calculation for SUM Delay calculation for CARRY
The Electrical effort H = 10pf/10pf (Output
capacitance / Input capacitance) = 1
Path logical effort G along path (N-W) =
1 2g g•∏ (where g1 and g2 are logical
effort of XOR gate) =4X4=16
The parasitic delay along path (N-W) P =
4+4=8
(The parasitic delay of XOR gate is 4)
The Branching effort B =1 [because all of
the fan-out’s along the path are one]
Number of stages N= 2 stages
The path Efforts F=GBH =(16)X1X1=16
The path delay
^
1/ N
D N PF= +
= 2X(16)1/2
+8 =16ps
The Electrical effort H = 10pf/10pf (Output
capacitance / Input capacitance) = 1
Path logical effort G along path (N-W) =
3 4g g•∏ (where g3 and g4 are logical
effort of NAND gate) =4/3X4/3 = 16/9
The parasitic delay along path (N-W)P
=2+2=4 (The parasitic delay of NAND gate
is 2)
The Branching effort B =1 [because all of the
fan-out’s along the path are one]
Number of stages N= 2 stages
The path Efforts F=GBH =(16/9)X1X1=16/9
The path delay
^
1/ N
D N PF= +
= 2X(16/9)1/2
+4 =6.6ps
Figure 2. Logical Delay Model for Full Adder Circuit.
So the total delay will be the sum of CARRY and SUM which is equal to 22.6ps. From this
observation the delay of the circuit vary with change in the input and output capacitance value.
4. SIMULATION AND PERFORMANCE ANALYSIS OF FULL ADDER
The proposed 20 different Boolean expressions (logic construction) are simulated using Tanner
EDA with BSIM3v3 250nm technology with supply voltage ranging from 1V to 2V in steps of
0.2V. All the full adders are simulated with multiple design corners (TT, FF, FS, and SS) to
verify that operation across variations in device characteristics and environment. The simulated
setup for optimized full adder’s (using XOR,MUX) test bed and its gate equivalent along with its
input/output waveform is shown in Figure ( 3 ). The test bed is supplied with a nominal voltage of
2V in steps of 0.2V and it is invoked with the technology library file Generic 025 and it is
specified with TT, FF, FS and SS conditions. The W/L ratios of both nMOS and pMOS
transistors are taken as 2.5/0.25µm. To establish an unbiased testing environment, the simulations
have been carried out using a comprehensive input signal pattern, which covers every possible
transition for a 1- bit full adder.

141
The frequencies have been chosen in the range from 10 to 200MHz and its input and output
capacitances are set to 10pf. The three inputs to the full adder are A, B, C and all the test vectors
are generated and have been fed into the adder cell. The cell delay has been measured from the
moment the inputs reach 50% of the voltage supply level to the moment the latest of the SUM
and CARRY signals reach the same voltage level. All transitions from an input combination to
another (total 8 patterns, 000, 001, 010, 011, 100, 101, 110, 111) have been tested, and the delay
at each transition has been measured. The average has been reported as the cell delay. The power
consumption is also measured for these input patterns and its average power has been reported in
Table 3.
Figure 3. A. Snap Shot of Full Adder with XOR and MUX
Figure 3.B . Test bed For Full Adder
The simulation results are shown in Table 3. The performance of all the full adders has been
analyzed in terms of delay, transistor count and power dissipation. It is observed that adder
designed with XOR and MUX has the least delay, transistor count and power dissipation when
compared to other combinations of gate. So the adder realized with MUX and XOR is considered

142
to be the optimized adder in terms of delay, transistor count and power dissipation. The second
optimized full adder is realized from XNOR, NOT and MUX.
Figure 3. C. Input/ Output Wave of Full Adder
The worst case full adder construction is not using NOR gate which occupies large transistor
count, dissipates large power and has longer delay. The other optimized solution for constructing
full adders are using NAND gates only, XOR, XNOR,MUX combination and XOR,
AND,OR,MUX combination. Figure (4) shows the simulation result of full adders in terms of
delay, transistor count (TC) and power dissipation (PWR-DISSP). The power-delay product all
the full adders is shown in Figure (5).
Figure 4. Simulation result of adders in terms Figure 5. Power delay product
delay, area and power

143
Table 3. Simulated Result for 20 different Full adders
5. FPGA IMPLEMENTATION
In this work the adder structures used are: Ripple Carry Adder, Carry Look-Ahead Adder, Carry
Save Adder, Carry Increment adder, Carry Select Adder, Carry Skip Adder. From section IV it is
observed that the optimized equation for implementing 1-bit full adder is using XOR and MUX.
So the primitive of this adder cell is implemented with multiplexer and this module is
incorporated with existing adder topologies. The target FPGA device chosen for the
implementation of these adders was Xilinx ISE 12.1 Spartan3E XC3S500-5FG320. This device
was chosen because the Spartan3E families of Field-Programmable Gate Arrays (FPGAs) are
specifically designed to meet the needs of high volume, cost-sensitive consumer electronic
applications. The Spartan-3E family builds on the success of the earlier Spartan-3 family by
increasing the amount of logic per I/O, significantly reducing the cost per logic cell. These
Spartan-3E FPGA enhancements, combined with advanced 90 nm process technology, deliver
more functionality and bandwidth. Each adder type was implemented with bit sizes of: 8, 16, 32,
64 bits. This variety of sizes will provide with more insight about the performance of each adder
in terms of area and delay as a function of size. Structural Gate level modeling using Verilog
HDL was used to model each adder. The Xilinx ISE Foundation version 12.1i software was used
for synthesis and implementation.
Table 4 contains the results obtained. The adder abbreviations used in the table are: RCA for
Ripple Carry Adder, CLA for Carry Look-Ahead Adder, CSA for Carry Save adder, CIA for
Carry Increment Adder, CSeA for Carry Select Adder and CSkA for Carry Skip Adder. In the
Table, delay is measured in nanoseconds (ns) while area is measured in Slice Look-Up Tables
(LUT) units which represent configurable logic units within the FPGA. In the Table E-A
represents the adder designed with normal expression for SUM and CARRY, Op-A represents
Full adder using
Avg
Delay
(ps)
Transistor
count
Avg
Power
Dissipation(µW)
XOR,AND, OR (1) 36.1 38 12.693
XOR,AND,OR(2) 33.1 30 9.813
XNOR,AND,OR 35 38 10.371
XNOR,AND,OR,NOT 34 32 9.833
XOR, AND 36.2 30 12.35
XOR, NAND, Negative OR 30.13 30 10.547
XNOR,NAND,NOT 29.5 34 8.752
XNOR, NAND 23.01 30 9.585
XOR,NAND 23.023 30 9.485
XOR, MUX 18.02 18 7.704
XNOR, MUX, NOT 19.01 20 7.769
XOR, XNOR, MUX 20.8 24 8.691
XOR, AND, OR, MUX 23.6 30 9.798
XNOR,AND, OR,MUX 26.12 30 9.058
NAND 20.2 36 10.795
NOR 40.1 51 22.25
XOR,NAND,NOR,NOT 32.2 38 10.583
XNOR,NAND,NOR,NOT 31.1 38 10.1
XNOR,NOR,NOT,OR 34.1 40 11.723
XOR,NOR,NOT,OR 35 42 12.487

144
Figure 6 Different Adder topologies

145
the adder topology implemented with optimized equations that are realized in terms of
multiplexers. It is noticed that delay, area and power delay product are less when compared to the
normal expression. Figure 6 shows different adder topologies
Table 4 Comparison and Results obtained for different adder topologies
6. SUMMARY
The proposed 20 different Boolean expressions (logic construction) are simulated using Tanner
EDA with BSIM3v3 250nm technology with supply voltage ranging from 1V to 2V in steps of
0.2V. It is observed that adder designed with XOR and MUX has the least delay, transistor count
and power dissipation when compared to other combinations of gate. So the adder realized with
MUX and XOR is considered to be the optimized adder in terms of delay, transistor count and
power dissipation. A new low-power, high-speed full adder cell is proposed using XOR and
MUX gates. Its performances have been analyzed and reported in section 4. This optimized adder
is designed with fully MUX based structure in FPGA using VERILOG HDL and this module is
incorporated in the existing adder topologies and its comparison is made. The target FPGA
device chosen for the implementation of these adders was Xilinx ISE 12.1 Spartan3E XC3S500-
5FG320. The comparison of delay, slice occupied, AT and its power dissipation is depicted in the
A
d
d
e
r
s
Slice
utilized
AT
Power
Dissipation
(mw) PD
Delay (ns) (Area)
bit
E-A
Op-
A E-A
Op-
A E-A Op-A E-A Op-A E-A Op-A
R 8 13.2 12.93 9 9 118.83 116.37 80.98 80.1 1069.18 1035.69
C 16 21.69 21.11 18 18 390.42 379.89 81.1 79.98 1759.06 1687.98
A 32 38.67 37.46 37 37 1430.6 1385.87 82.3 78.5 3182.13 2940.3
64 72.62 70.16 74 74 5373.6 5191.77 85.67 80.2 6221.01 5626.75
C 8 13.2 12.93 9 9 118.83 116.37 78.91 77.5 1041.85 1002.08
L 16 21.69 21.11 18 18 390.42 379.89 80.98 78.2 1756.46 1650.41
A 32 38.67 37.46 37 37 1430.6 1385.87 81.2 79.1 3139.6 2962.77
64 72.62 70.16 74 74 5373.6 5191.77 82.3 79.3 5976.3 5563.61
C 8 12.06 11.1 14 13 168.81 144.3 85.23 80.98 1027.7 898.878
S 16 20.02 19.8 27 23 540.49 455.4 87.1 82.3 1743.57 1629.54
A 32 34.94 32.12 55 52 1921.5 1670.24 88.45 82.3 3090.18 2643.48
64 56.8 54.23 110 106 6247.9 5748.38 90.12 84.01 5118.73 4555.86
C 8 12.21 11.9 12 11 146.56 130.9 85.23 80.18 1040.91 954.142
I 16 16.57 14.32 24 22 397.66 315.04 87.2 82.03 1444.82 1174.67
A 32 27.45 25.67 49 47 1345.1 1206.49 88.23 82.03 2422.09 2105.71
64 47.2 45.21 100 98 4719.7 4430.58 90.1 84.12 4252.45 3803.07
C 8 12.6 10.11 15 11 188.94 111.188 78.91 77.5 993.95 783.37
S 16 21 15.75 32 30 672.1 472.56 80.98 78.2 1700.82 1231.81
E 32 37.93 24.36 67 61 2541 1486.2 81.2 79.1 3079.51 1927.19
A 64 71.77 31.41 135 125 9689 8790.12 82.3 79.3 5906.67 2490.65
C
S
K
A
8 12.78 11.15 13 13 166.17 144.885 78.91 77.5 1008.63 863.738
16 23.27 14.52 23 22 535.28 319.352 80.98 78.2 1884.65 1135.15
32 40.15 23.26 45 44 1806.9 1023.22 81.2 79.1 3260.5 1839.47
64 49.22 35.73 79 78 3888.5 2787.1 82.3 79.3 4050.97 2833.55

146
Figure (7). From this analysis it is found that for all the adder topologies the delay is less when
compare to the existing adder with normal equation ( S U M A B C= ⊕ ⊕ and
CARRY AB AC BC= + + ) and it is also observed that the delay for RCA and CLA are the same and
its distribution is shown in the graph (Figure 7a). In case of slice utilized there is no change
occurs for RCA and CLA hence its distribution is shown as single red line in the chart (Figure
7b). From AT chart (Figure 7c) it is noticed that the AT value is large for 64 bit carry select
adders and adders like ripple carry adder, carry look ahead adder and carry increment adder have
less AT Value. From PD distribution (Figure 7d) less power dissipation occurs for carry
increment and ripple carry adders, maximum dissipation occurs for carry save and carry skip
adders. According to the presented results, the adder topology which has the best compromise
between area, delay and power dissipation are carry look-ahead and carry increment adders and
they are suitable for high performance and low-power circuits. The fastest adders are carry select
and carry save adders with the penalty of area. The simplest adder topologies that are suitable for
low power applications are ripple carry adder, carry skip and carry bypass adder with least gate
count and maximum delay.
Figure 7. A. Delay Chart
Figure 7. B. Slices utilized Chart

147
Figure 7.C. AT Value Chart
Figure 7. D. PT Value Chart
7. CONCLUSION
An extensive performance analysis of 1-bit full-adder cells has been presented. Technology
independent logic optimization is used to design 1-bit full adder with 20 different Boolean
expressions and its performance was analyzed in terms of transistor count, delay and power
dissipation using Tanner EDA with TSMC MOSIS 250nm technology. From this analysis XOR
and MUX based expression provides low transistor count, minimum delay and minimum power
dissipation when compared to other logic equations. The second optimized full adder can be
realized using XNOR, NOT and MUX. The other optimized solution for constructing full adders
are using NAND gates only, XOR, XNOR,MUX combination and XOR, AND,OR,MUX
combination. The worst case full adder construction is not using NOR gate which occupies large
transistor count, dissipates large power and has longer delay. Logical effort delay model to
estimate the parasitic delay is also presented. Using the optimized expression the primitive adder
cell is implemented with multiplexer and this module is incorporated with existing adder
topologies like ripple carry adder, carry look-ahead adder, carry skip adder, carry select adder,
carry increment adder and carry save adder and its performance is analyzed in terms of area
(slices used) and maximum combinational path delay as a function of size. The target FPGA
device chosen for the implementation of these adders was Xilinx ISE 12.1 Spartan3E XC3S500-
5FG320. The comparison and its simulation results have been presented. Based on the
comparison it is observed that number of slices occupied, power dissipation and delay are less
using the optimized expression. The work presented in this paper gives more insight and deeper
understanding of constituting modules of the adder cell to help the designers in making their
choices.

148
REFERENCES
[1] Shih-Chieh Chang, Lukas P.P.P. van Ginneken, “Circuit Optimization by Rewiring”, IEEE
Transaction on Computers, Vol. 48, No. 9, September 1999.
[2] Oh-Hyeong Kwon, “A Boolean Extraction Technique For Multiple-Level Logic Optimization” IEEE
2003
[3] R.Uma, “4-Bit Fast Adder Design: Topology and Layout with Self-Resetting Logic for Low Power
VLSI Circuits”, International Journal of Advanced Engineering Sciences and Technology, Vol No. 7,
Issue No. 2, 197 – 205, 2011.
[4] Padma Devi, Ashima Girdher and Balwinder Singh, “Improved Carry Select Adder with Reduced
Area and Low Power Consumption”, International Journal of Computer Application,Vol 3.No.4, June
2010
[5] Shrirang K. Karandikar and Sachin S. Sapatnekar,” Fast Comparisons of Circuit Implementations”,
IEEE Transaction on Very Large Scale Integration (VLSI) Systems, Vol. 13, No. 12, December 2005
[6] Sutherland, B. Sproull, D. Harris, “Logical Effort: Designing Fast CMOS Circuits,” Morgan
Kaufmann Publisher, c1999.
[7] R.uma, Vidya Vijayan, M. Mohanapriya, Sharon Paul,” Area, Delay and Power Comparison of
Adder Topologies”, International Journal of VLSI and Communication Systems, 2012.
[8] G.Shyam Kishore, “A Novel Full Adder with High Speed Low Area”, 2nd National Conference on
Information and Communication Technology (NCICT) 2011 Proceedings published in International
Journal of Computer Applications® (IJCA).
[9] Mariano Aguirre-Hernandez and Monico Linares-Aranda, “CMOS Full-Adders for Energy-Efﬁcient
Arithmetic Applications”, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Vol.
19, No. 4, April 2011.
[10] Pudi. V, Sridhara., K, “Low Complexity Design of Ripple Carry and Brent Kung Adders in
CA”,Nanotechnology, IEEE transactions on,Vol.11,Issue.1,pp.105-119,2012.
[11] Shubin.V.V, “Analysis and Comparison of Ripple Carry Full Adders by Speed”, Micro/Nano
Technologies and Electron Devices (EDM),2010, International Conference and Seminar on, pp.132-
135,2010.
[12] Sreehari Veeramachaneni, M.B. Srinivas, “New Improved 1-Bit Full Adder Cells”, IEEE, 2008.
[13] Ivan E. Sutherland, Bob F. Sproull, David L. Harris “Logical Effort: Designing Fast CMOS Circuits”,
Morgan Kaufmann Publishers, Inc. 1998.
[14] R. F. Sproull and I. E. Sutherland. “Theory of Logical Effort: Designing for Speed on the Back of an
Envelope”. In IEEE Adv. Res. in VLSI, 1–16, 1991.
[15] Jian-Fei Jiang; Zhi-Gang Mao; Wei-Feng He; Qin Wang, “A New Full Adder Design for Tree
Structured Arithmetic Circuits”, Computer Engineering and Technology(ICCET),2010,2nd
International Conference on,Vol.4,pp.V4-246-V4- 249,2010.
[16] Shubhajit Roy Chowdhury, Aritra Banerjee, Aniruddha Roy, Hiranmay Saha, “A high Speed 8
Transistor Full Adder Design using Novel 3 Transistor XOR Gates”, International Journal of
Electrical and Computer Engineering 3:12 2008.
[15] Romana Yousuf and Najeeb-ud-din, “Synthesis of Carry Select Adder in 65 nm FPGA”, IEEE.
[16] D. Patel, P. G. Parate, P. S. Patil, and S. Subbaraman, ―ASIC implementation of 1-bit full
adder,ǁ in Proc. 1st Int. Conf. Emerging TrendsEng. Technol., Jul. 2008, pp. 463–467.
[17] Y. Sunil Gavaskar Reddy and V.V.G.S.Rajendra Prasad, “Power Comparison of CMOS and
Adiabatic Full Adder Circuits”, International Journal of VLSI design & Communication Systems
(VLSICS) Vol.2, No.3, September 2011

149
Authors
R.Uma received her B.E. (EEE) degree from Bharathiyar University Coimbatore
in the year 1998, Post graduated in M.E (VLSI Design) from Anna
University Chennai in the year 2004. Currently she has been working as Assistant
Professor in Electronics and Communication Engineering, Rajiv Gandhi College
of Engineering and Technology, Puducherry. She authored books on VLSI Design.
She has published several papers on national and International journal and
conferences. She is the guest faculty for Pondicherry University for M.Tech
Electronics. She has received the best teacher award for the year 2006 and 2007.
Her research interests are Analog VLSI Design, Low power VLSI Design,
Testing of VLSI Circuits, Embedded systems and Image processing. She is a member of ISTE. Perusing her
Ph.D. from Pondicherry University in the Department of Computer Science.
Dr. P. Dhavachelvan is working as a Professor in the Department of Computer
Science, Pondicherry University, India. He obtained his B.E. in the field of
Electrical and Electronics Engineering from University of Madras, India. He
pursued his M.E. and Ph.D. in the field of Computer Science and Engineering from
Anna University, Chennai, India. He has about 15 years of experience as an
academician and his research areas include Software Engineering & Standards and
Web Service Computing. In his credit, he has more than 125 research papers
published in reputed International and National Journals and Conferences. He also
obtained Patents and proposed Standards in the domain of Software Engineering.

LOGIC OPTIMIZATION USING TECHNOLOGY INDEPENDENT MUX BASED ADDERS IN FPGA

Recommended

More Related Content

What's hot (18)

Similar to LOGIC OPTIMIZATION USING TECHNOLOGY INDEPENDENT MUX BASED ADDERS IN FPGA (20)

Recently uploaded (20)

LOGIC OPTIMIZATION USING TECHNOLOGY INDEPENDENT MUX BASED ADDERS IN FPGA