Crossbar Array of Artificial Synapses Based on Ferroelectric Diodes
Laura Seufert, Morteza Hassanpour Amiri, Paschalis Gkoupidenis, Kamal Asadi
Abstract
Abstract Two terminal devices that exhibit resistance switching in response to an external voltage are interesting for neuromorphic computing applications. Owing to its simple device structure, a crossbar array of two‐terminal resistance switching devices is highly desired for application as artificial neural network weights. Here, ferroelectric diodes that show resistance switching in their forward bias are presented. The resistance can be set to a high‐ and a low‐resistance state or any state between these limits. It is demonstrated that the ferroelectric diodes can function as an artificial synapse. An array of the ferroelectric diodes with two bit and two row lines (2 × 2) is demonstrated. The resistance of every bit is independently tuned, and spike‐time‐dependent plasticity is shown for the array.