| Abstract | As transistor dimensions approach atomic scale,
quantum-mechanical effects such as tunneling and spin become important
ingredients in accurate performance models of integrated
circuits. Theoretical work in terms of such models suggests that
power-density constraints may eventually require a departure from
common practices of representing logic 0s and 1s by charges, voltages
or currents. Instead, nuclear and electron spins are proposed as
primary careers of stationary information, e.g., in the
well-publicized demonstration by IBM in 2000, and photon polarizations
can transport quantum information over great distances, acting as
quantum bits. However, the algebra of quantum bits is radically
different from the Boolean algebra that describes modern digital
electronics, while such states are succeptible to frequent and unusual
types of errors. On the positive side, quantum communication promises
an unparalleled level of security and some quantum algorithms solve
other-wise intractable problems in polynomial time. Despite many
potential applications and several active start-ups in the field, the
main obstacle to further progress in quantum information processing is
complexity. This is where design automation can lend a helping hand. |