Adaptive computation by interacting quantum dots

We present a theoretical study and discussion of computationally useful nanoelectronic circuits which use adaptive control methods both to achieve the circuit function and to compensate for unpredictable nonuniformities in the circuit environment. In the regime where the scaling of conventional digital electronics breaks down, nanoelectronic circuitry will be required to perform robustly in the presence of inevitable device–device interactions, sensitivity to circuit parameters of quantum devices, and deviations from ideal circuit design. To examine the role of adaption in addressing these issues, we focus on a specific class of scaleable circuit architectures composed of Coulombically interacting polarizable anisotropic quantum dots which include input polarization dots, output polarization dots, and an array of processing dots. We implement the adaptive control of these circuits by assuming that particular features of the processing dots such as energy barriers, charge, shape, or orientation can be experimentally modified. A method of adaptive feedback is used to modify the processing dots and produce desired correlations between the input and output dot polarizations as computed by the circuit. A variational quantum Monte Carlo method has been used to simulate the many-body response of model GaAs dot circuits in which the mutual orientation of the dots is adapted to successfully achieve different desired patterns of correlation. We demonstrate the robustness of the adaptive circuits for circuit nonuniformities and for sensitivity to circuit parameters due to quantum effects.