Information Technology Reference
In-Depth Information
Fig. 1. Nano-scale magnets with varying sizes and aspect ratios. Single-domain behavior is
observed if the magnets are sufficiently small and narrow. (Source: A. Imre, Ph.D. dissertation,
University of Notre Dame, 2005 [ 5 ].)
flux closure inside, without strong coupling to the exterior. Such single-domain
magnets, with typical size scales of tens to hundreds of nanometers, form the physical
basis for NML.
3
From Quantum-Dot Cellular Automata to Nanomagnet
Logic
Our current work on nanomagnet logic grew out of our previous work on Quantum-
Dot Cellular Automata (QCA), which was an attempt to base computation on phys-
ically-interacting cellular arrays of quantum dots occupied by a few electrons [ 6 ].
Instead of wires, neighboring devices interact through direct Coulomb interactions
between electrons on neighboring quantum dots. We have shown that such physical
interactions in appropriately structured arrays of quantum dots can also be used to
realize logic gates. Electronic implementations of QCA proved difficult due to tech-
nological limitations of quantum-dot fabrication (such as size variations) and elec-
tronic stray charges. For a review of electronic QCA, see Refs. [ 7 , 8 ].
Nanomagnet logic can be viewed as a magnetic implementation of QCA.
(In earlier publications, we used the term magnetic QCA (MQCA), but we now prefer
to use NML in order to avoid confusion with quantum dots.) Early theoretical work on
magnetic QCA is given in the Ph.D. Dissertation of György Csaba [ 9 ]. These simu-
lations demonstrated the feasibility of using field-coupled single-domain nanomagnets
for realizing basic logic functionality [ 10 - 12 ]. In subsequent experimental work
stimulated by these simulations, and which constituted the Ph.D. Dissertation of
Alexandra Imre [ 5 ], we first demonstrated magnetic wires formed by chains of near-by
magnetic islands. Since the individual dots have elongated shapes, there are two basic
types of wire arrangements. In one type, the magnets are lined up side-by-side and, as
one's intuition would tell from two bar magnets next to each other with their long
sides, the individual magnets prefer to be magnetized in the opposite direction; we call
Search WWH ::




Custom Search