International Journal of Applied Engineering Research
ISSN 0973-4562 Volume 9, Number 21 (2014) pp. 8889-8900
© Research India Publications
http://www.ripublication.com
Area and Power Optimized Wallace Tree Multiplier using
Power Gating Technique: A Transistor Level Design
Anitha R.
1
, S Jakir Hussain
2
, Sarat Kumar Sahoo
3
1
School of Electronics Engineering, ranitha@vit.ac.in
2
School of Electronics Engineering, zakirhussain1218@gmail.com
3
School of Electrical Engineering, sksahoo@vit.ac.in
VIT UNIVERSITY Vellore, INDIA\
Abstract
In this work we have designed the Wallace tree multiplier at transistor level. In
most of the multipliers XOR gate is a key component. The optimization is
done at gate level by using the 6T XOR gate instead of conventional CMOS
gate. A power gating technique is also used to reduce the leakage power in the
design. The simulation was carried out in CADENCE virtuoso with 90nm
technology. And the result showed a reduction in area and power of the
multiplier.
Index Terms— Wallace tree, XOR gate, 6T XOR gate, power gating, 90nm
technology, Cadence
Introduction
The greater emphasis now a day in IC design is performance and miniaturization with
reduced power and area. The growing trend towards portable computing and
communication systems has made engineers to design systems with reduced area and
power. Multiplier is the most important block in most of the digital systems like DSP
Processors, microprocessors, etc,. With the advances in technology may researchers
have done lot of research to design multipliers with the following specification such
as high speed, low power and low area or less delay. However area and speed are
inversely proportional to each other i.e. larger the area leads to improving speed.
Array multipliers are most commonly used multipliers. The main drawback of the
array multiplier is that high power consumption and more digital gates resulting in the
increased area of the multiplier. There are various design of the array multiplier.
Wallace tree multiplier has been designed with the compressors so as to reduce the
latency in [1]. In work carried out in [1] has mainly concentrated on the delay of the