Since the voltage on node n1 for the various threshold gates described herein is closer to rail-to-rail voltage compared to the case when linear input capacitors are used, subsequent driver circuitry 1901 can be removed. As such, the input capacitors with nonlinear polar material can drive another capacitive input circuit directly. Here, closer to rail-to-rail voltage on node n1 using nonlinear polar material based capacitors (e.g., ferroelectric or paraelectric capacitors) implies that the static leakage in the subsequent driver 1901 is reduced compared to the case when voltage on n1 is not close to rail-to-rail voltage. When linear capacitors are used, a voltage divider is formed on node n1 based on the number of capacitors and their logic inputs. Such a voltage divider results in non-rail-to-rail voltage on node n1 that results in static leakage in the subsequent driver 1901. When nonlinear capacitors are used, the voltage divider is not a linear voltage divider. This results in a much closer rail-to-rail voltage on n1 which reduces static leakage in the subsequent driver 1901. The higher the nonlinearity, the closer the voltage on node n1 is rail-to-rail. Nonlinear capacitors as shown in various embodiments allow the logic gate to have more inputs compared to the case when linear capacitors are used while keeping the leakage through driver 1901 low.
In various embodiments, 5-input configurable threshold gate 1900 can be configured as a stack and fold configuration as described with reference to