Information Technology Reference
In-Depth Information
NanoMagnet Logic:
An Architectural Level Overview
B
Marco Vacca 1 , Mariagrazia Graziano 1(
) , Juanchi Wang 1 ,
Fabrizio Cairo 1 , Giovanni Causapruno 1 , Gianvito Urgese 2 ,
Andrea Biroli 1 , and Maurizio Zamboni 1
1 Dipartimento di elettronica e telecomunicazioni, Politecnico di Torino, Turin, Italy
mariagrazia.graziano@polito.it
2 Dipartimento di Ingegneria informatica e dei sistemi,
Politecnico di Torino, Turin, Italy
Abstract. In recent years Field-Coupled devices, like Quantum dot
Cellular Automata, are gaining an ever increasing attention from the
scientific community. The computational paradigm beyond this device
topology is based on the interaction among neighbor cells to propa-
gate information through circuits. Among the various implementations of
this theoretical principle, NanoMagnet Logic (NML) is one of the most
studied. The reason lies to some interesting features, like the possibility
to combine memory and logic in the same device and the possible low
power consumption. Since the working principle of Field-Coupled devices
is completely different from CMOS technology, it is important to under-
stand all the implications that this new computational paradigm has on
complex circuit architectures.
In this chapter we deeply analyze the major issues encountered in the
design of complex circuits using Field-Coupled devices. Problems are
analyzed and techniques to solve them and to improve performance are
presented. Finally, a realistic analysis of the applications best suited for
this technology is presented. While the analysis is performed using Nano-
Magnet Logic as target, the results can be applied to all Field-Coupled
devices. This chapter therefore supplies researchers and designers with
the essential guidelines necessary to design complex circuits using Nano-
Magnet Logic and, more in general, Field-Coupled devices.
1
Introduction
With the conception of Quantum dot Cellular Automata (QCA) [ 1 ] technology,
a completely new computational paradigm to process and propagate informa-
tion was presented to the scientific community. Signals were no more represented
by voltage or current levels, but by charge configurations. Circuits are made by
many identical cells and information processing is driven by electrostatic interac-
tion among neighbor cells [ 2 ]. Different practical implementations of this princi-
ple were proposed, the two most promising are Molecular QCA [ 3 ] and Magnetic
QCA also called NanoMagnet Logic [ 4 ]. Molecular QCA [ 5 - 8 ] use complex mole-
cules as basic cells. The main interest for this QCA implementation resides in
 
Search WWH ::




Custom Search