Development of an Algorithm for 16-Bit WTM

IOSR Journal of Electronics and Communication Engineering (IOSR-JECE)
e-ISSN: 2278-2834,p- ISSN: 2278-8735.Volume 10, Issue 1, Ver. III (Jan - Feb. 2015), PP 79-86
www.iosrjournals.org
DOI: 10.9790/2834-10137986 www.iosrjournals.org 79 | Page
Development of an Algorithm for 16-Bit WTM
1
Sravanthi.kantamaneni, Asst Professor, 2
Dr.V.V.K.D.V.Prasad, Professor,
3
Veera Vasantha Rao.Battula, Asst. Professor
Abstract: Binary Multipliers plays an important role in digital circuits. There are many methods for
generating a Simple binary multiplication and some of them are like Ripple carry array multipliers, Row adder
tree multipliers, Partial product LUT multipliers, Wallace trees, Booth recoding etc.,. Our project mainly
concentrates on 8x8 Wallace tree multiplier. It uses a famous Wallace tree structure which is an implementation
of an adder tree designed for minimum propagation delay. Rather than completely adding the partial products
in pairs like the ripple adder tree does, the Wallace tree sums up all the bits of the same weights in a merged
tree. Usually full adders are used, so that 3 equally weighted bits are combined to produce two bits: one (the
carry) with weight of n+1 and the other (the sum) with weight n. Each layer of the tree therefore reduces the
number of vectors by a factor of 3:2. A conventional adder is used to combine these to obtain the final product.
The benefits of the Wallace tree is that there are only O(logn) reduction layers, and each layer has O(1)
propagation delay. As making the partial products is O(1) and the final addition is O(logn), the multiplication is
only O(logn), not much slower than addition (however, much more expensive in the gate count). Naively adding
partial products with regular adders would require O(log2
n) time. Our project is to develop 8 x 8 Wallace tree
multiplier using VHDL and will be simulated with the help of XILINX simulator and verified on Spartan-3E
FPGA circuit board.
I. Introduction
Recent advancements in mobile computing and multimedia applications demand for high performance
and low-power consuming VLSI (very large scale integrated circuit) Digital Signal Processing (DSP) systems.
One of the most important components of DSP systems is a multiplier. Multiplication is basically shift and add
operation. Usually in a DSP system, multiplier units consume large amount of power and cause most of the
delay compared to other units like adders. Depending on size of the inputs (2 X 2 bit, 4 X 4, 8 X 8 etc.,) the
number of steps a normal binary multiplier takes to compute the product increases drastically. Larger the steps
of calculation larger will be the delay as well as the power consumption. Also area occupied by the multiplier on
a FPGA (Field Programmable Gate Array) increases. Hence various algorithms have been developed in order to
achieve lesser complexity in computation involving minimum calculation steps, which in turn can reduce delay,
power and area constraints of multipliers.
The Wallace tree has three computation steps:
1. Generation of Partial products – multiplying each bit of one binary input with every bit of the other binary
input. If each input has n-bits the result of this step will give us n2
number of binary bits called „Partial
products‟ distributed in n-rows and 2n-columns. This step is very same as what we do to multiply two
numbers by hand.
2. Reduction of partial products – the partial products are to be added according to their place values (or
„weights‟) using half adders and full adders until only two rows of partial products are left.
3. Last stage addition – remaining two rows will be added using a conventional adder to get final result of
multiplication.

Wallace tree multiplier reduction stages for 8X8 multiplication [1]

Final Result Bits
Improvements to the Algorithm:
First stage:
It is evident that every multiplication in Wallace tree algorithm is done in three logical stages. They are
the partial product generation, the reduction stages and the last stage addition using a conventional adder. If we
recall the description for Partial product generation, the first step, we are doing it by use of AND gates. If two
8 bit numbers say A and B are to be multiplied, the algorithm starts at the least significant bits (LSBs) of A and
B. LSB of B say b0 will be AND with all the eight bits of A from a0 to a7. This gives us a single row of partial

products. The second row of PPs will be generated when all bits of A will be AND with b1. Similarly, every row
will be formed due to AND operations.
Rows of partial products of 8 bit multiplication
Every partial product of first row has b0. Every second row element has b1. Similarly every partial
product of nth
row will have the common AND input bn-1. So, we can write mathematically the first row as (a7 a6
a5 a4 a3 a2 a1)*b0. Any nth
row can be written as (a7 a6 a5 a4 a3 a2 a1)*bn-1.
We know that all the above bits are binary digits i.e. either 0 or 1. Hence two possibilities exist.
If b0 = 0:
Then the first row (a7 a6 a5 a4 a3 a2 a1)*b0 will be equal to
(a7 a6 a5 a4 a3 a2 a1)*0 = (0 0 0 0 0 0 0 0)
If b0 =1:
Then the first row (a7 a6 a5 a4 a3 a2 a1)*b0 will be equal to
(a7 a6 a5 a4 a3 a2 a1)*1 = (a7 a6 a5 a4 a3 a2 a1) = A
Hence any row will be equal to the Multiplicand A or it will be a row full of Zeros.
Keeping the above fact in mind we can use another way to generate partial products for our need
without using n2
number of AND gates (n is the size of inputs). This method is given below. Any nth
row will be
0 or A based on the value of common input bn-1 of that row.
Row „n‟ = 0, if bn-1=0
Row „n‟ =A, if bn-1=1
Consider the following example to comprehend this new logic in a better way.
Let A= 11111111 and B=10011011. A is multiplicand and B is multiplier as usual.
The rows of partial products for the multiplication A X B are:
Row 1 = b0 AND (11111111) = 1 AND (11111111) =11111111 =A
Row 2 = b1 AND (11111111) = 0 AND (11111111) =11111111 =A
Row 3 = b2 AND (11111111) = 0 AND (11111111) =00000000 =O
Row 4 = b3 AND (11111111) = 1 AND (11111111) =11111111 =A
Row 5 = b4 AND (11111111) = 1 AND (11111111) =11111111 =A
Row 6 = b5 AND (11111111) = 0 AND (11111111) =00000000 =O
Row 7 = b6 AND (11111111) = 0 AND (11111111) =00000000 =O
Row 8 = b7 AND (11111111) = 1 AND (11111111) =11111111 =A
(Or)
Simply we can write:
Row 1 = A since b0= 1
Row 2 = O since b1= 0

Advantage:
In case of previous method for a n-bit multiplier, the first stage will require a total of n2
number of
AND gates. But with the new method the task of partial product generation will be done by just n number of
steps instead of n2
steps. It is important to note that while describing the multiplier in VHDL code, each of the n2
AND operations have to be written manually. For a 32 bit- multiplier it requires 1024 steps to be written for
simple AND operations. Instead, with the new modification, only 32 steps are to be written which will save a lot
of energy and time to the design engineer during development of the code. So, we have adopted the latter
method in designing the WTM system.
Representation of signals:
As described in an algorithm, we have to use different variables to indicate the partial products in
different levels of reduction stages. For example, we used a, b in first stage, then P0, P1 in the next and later S,
C, M, N etc. in the figures of chapter 2. These variables are of our choice and we must make sure that PPs in
different levels of reduction do not have the same representation. It means that designer has to make a note of
which variables he is using in what stage of reduction, clearly and without confusion.
There is another sound drawback of representing variables (or „signals‟ with respect to VHDL coding)
using normal alphabets like A, B, M or N etc. Let us assume that we have come across a signal N2 while
verifying the design. We cannot readily identify which reduction stage this signal N2 belongs to. We must go
through the code once again form start and locate where N2 has its origin. Imagine a 32 bit multiplier which will
have a very large number of such signals. To go through the code every other time to know about a signal, it is a
tremendous burden for the designer. So, it is of high importance that we have a proper representation scheme for
signals or variables. We must be able to identify from the name of a signal or variable several aspects. They are:
1. The reduction stage to which it belongs
2. The column or the weight of the partial product
3. Whether it is a SUM bit or CARRY bit
4. If more than one sum and carry bits are present in the column, then position of that bit in the column.
Scheme of representation of signals
To satisfy the above four requirements, we adopt the above representation scheme.

Ri reduction stage number „i‟ ; Eg: R1, R2, R4 etc.
Sj sum bit of column „j+1‟
Cj carry bit of column „j+1‟Weight of partial product = 2j
k place of the signal in the column.
If there are 4 sum bits in the column k takes the values of 1, 2, 3 and 4.
Always the sum bits are taken first and the carry bits are taken next to sums. Reverse will also give the same
answer, but to avoid confusion sums are given first priority in any column. Let us consider a column having 4
sums and 3 carries. Let all the bits belong to column number 6 (j=5) of 3rd
reduction stage. It will be represented
as follows.
R3S5_1
R3S5_2
R3S5_3
R3S5_4
R3C5_1
R3C5_2
R3C5_3
A column of signals of 3rd
reduction stage, 5th
column
Designing, Synthesis and Results of WTM for the Spartan 3E family FPGA chip
Simulated output of WTM
Maximum Combinational Path Delay
S. No
Size of
multiplier
Maximum combinational path delay (Nano seconds)
1 4 10.426
2 6 11.921
3 8 13.924
4. 10 14.775
5. 12 14.798
6. 14 16.168
7. 16 16.476

Comparison of Multipliers w.r.t Delay (ns)
Comparison of Multipliers in terms of Area
Comparison of Multipliers in terms of PDP
Comparison of Multipliers in terms of ADP
Size of the multiplier Vs No. of reduction stages
13.924
18.641
26.642
0
20
40
Delay(ns) 8*8 Multiplier
WTM
CSHM
Booth Multiplier
146
169 180
0
100
200
Area(LUT)
8*8 Multiplier
WTM
CSHM
Booth Multiplier
0.724
0.969
1.385
0
1
2
PDP
8*8 Multiplier
WTM
CSHM
Booth Multiplier
8*8 Multiplier
WTM 0.271
0.271 0.2353
0.47950
1
ADP
8*8 Multiplier
WTM CSHM Booth Multiplier
S. No Size of
multiplier
No. of Reduction
Stages Including last stage
1 4 3 4
2 8 4 5
3 16 6 7

Features of WTM
Conclusion and Discussion
It can be concluded that Wallace tree multiplier is superior in all respects like Delay, Area and speed.
However array multiplier requires more power consumption and gives optimum number of components
required, but it can provides a better delay .If we utilize this multiplier as a module of real time applications like
FIR Filter ,it can pump up the Filtering action. Further the work can be extended for optimization of said
multiplier to improve speed or to minimize the Power Consumption.
References
[1]. C. S. Wallace, A Suggestion for a Fast Multiplier, IEEE Transactions on Electronic Computers, February 1964, EC-13:14–17.
[2]. Vijaya Prakash A. M, Dr. MGR, K. S. Gurumurthy, A Novel VLSI Architecture for Low power FIR Filter, International Journal of
Advanced Engineering & Application, January 2011, PP 218 - 224.
[3]. Gary W. Bewick, Fast multiplication algorithms and implementation, The Department Of Electrical Engineering and The
Committee on Graduate studies of STANFORD UNIVERSITY, February 1994. PP 8 - 16.
[4]. J. Bhasker, AVHDL Primer, Third Edition, Pearson Education, 2007, PP-21 to 50, 88 to 101
[5]. John F. Wakerly, Digital Design Principle and Practices, fourth edition, Prentice Hall Pearson Education, 2009, PP 235-250, PP
786-795
[6]. http://en.wikipedia.org/wiki/Wallace_tree, http://en.wikipedia.org/wiki/FPGA.
[7]. Multiplication in FPGA‟s “The performance FPGA DESIGN specialist”
Parameter Used Available Pre Layout Values (or)
Ratio
Number Of Slices 96 371 2448 15%
Number of 4-input LUTs
178
647 4896 13%
Number Of Bonded
Input 32
64 158 40%
Number Of Bonded
Output 32
64 158 40%
Delay(ns) 16.476
Slice Utilization Ratio 100
BRAM Utilization Ratio 100

Development of an Algorithm for 16-Bit WTM

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Development of an Algorithm for 16-Bit WTM (20)

More from IOSR Journals (20)

Recently uploaded (20)

Development of an Algorithm for 16-Bit WTM