Modular Dissipation Analysis for QCA - Field-Coupled Nanocomputing: Paradigms, Progress, and Perspectives - page 353

Modular Dissipation Analysis for QCA

B Ilke Ercan and Neal G. Anderson (

)

Department of Electrical and Computer Engineering,
University of Massachusetts Amherst, Amherst, MA, USA
{ iercan,anderson } @ecs.umass.edu

Abstract. A modular approach for determination of lower bounds on
heat dissipation in clocked quantum-cellular automata (QCA) circuits
is proposed, and its application is illustrated. This approach, which is
based on a methodology developed previously for determining dissipa-
tion bounds in nanocomputing technologies, simplifies analysis of clocked
QCA circuits that are designed according to specified design rules. Fun-
damental lower bounds on the dissipative costs of irreversible informa-
tion loss for a (generally large and complex) QCA circuit are obtained
in the modular approach by (i) decomposing the circuit into smaller
zones, (ii) obtaining dissipation bounds for the individual circuit zones,
and (iii) combining results from the individual-zone analyses into a sin-
gle bound for the full circuit. The decomposition strategy is specifically
designed to enable this analytical simplification while ensuring that the
consequences of intercellular interactions across zone boundaries - inter-
actions that determine the reversibility of local information loss in indi-
vidual zone - are fully preserved and properly captured in the modular
analysis. Application of this approach to dissipation analysis of a QCA
half adder is illustrated, and prospects of using the modular approach
for automation of QCA dissipation analyses is briefly discussed.

· ·

Keywords: Nanocomputing QCA Heat dissipation

1 Introduction

Field-coupled nanocomputing (FCN) paradigms - both electronic and magnetic
- offer the possibility of computation at energy eciencies far superior to what
CMOS will ultimately provide. Logical operations are driven by interactions
between primitive computing elements that irreversibly discard energy and infor-
mation without exchanging mass. This obviates the energetically costly require-
ment for constant “opening” of the elements to one another and to particle
reservoirs, as is required in CMOS to transfer information and maintain the
computational “working substance.”
However, any computing paradigm that requires irreversible loss of informa-
tion - e.g. during implementation of logical transformations and erasing and/or
overwriting information remaining from previous computations - is necessarily
dissipative [ 1 , 2 ]. FCN circuits are no exception in this regard. Heat dissipation