Nonlinear Dynamics-Based Adaptive Hardware.

In this paper we briefly review some recent results in the field of nonlinear dynamics-based computing and how it can be utilized to implement flexible, reprogrammable hardware. Then we propose that such internal flexibility can be utilized to compensate for the negative effects of both internal and external changes. More specifically, we report results of an experimental study that we performed on an integrated circuit fabricated for nonlinear dynamics-based computing, where we overheated the circuit to a point that its operation and performance changed, but then we reprogramed the circuit so that the overheated circuit could again perform the same operations reliably. This inherent flexibility and reprogrammability opens the door to implementing adaptive circuits and systems that can withstand both variable environments and adverse conditions.