The remanent-polarization states of ferroelectric capacitors have long been of interest for nonvolatile storage of digital data. A simple memory array can be formed by perpendicular rows and columns of conductors with a ferroelectric capacitor at each crosspoint. In the simplest geometry, these conductors also are the capacitor electrodes on either side of a ferroelectric layer. Each capacitor represents one bit of digital data. If a voltage of V0 is required to write a capacitor, then data can be written to a specific bit by applying V0/2 and –V0/2 to the appropriate row and column. Such ferroelectric memory arrays were constructed as early as the 1950s. However, due to the breath of the ferroelectric transition, they were not successful because during the write of a specific capacitor, the V0/2 voltage signals disturb the polarization in other capacitors. This is in contrast to the successful use of the analogous ferromagnetic crosspoint arrays, which are familiar as the core memories of early computer technology.