Abstract
A highly scalable synapse device based on a junctionless (JL) ferroelectric (FE) FinFET is presented for neuromorphic applications. The synaptic behaviors of the JL metal-ferroelectric-insulator-silicon FinFET were experimentally demonstrated after verifying the ferroelectric characteristics of the HfZrOX (HZO) film using a metal-ferroelectric-metal capacitor. The fabricated synapse showed distinguishable polarization switching behaviors with gradually controllable channel conductance. From neural network simulations using the proposed JL FE FinFET as synapses, the pattern recognition accuracy for hand-written digits was validated to be approximately 80% for neuromorphic applications.
Original language | English |
---|---|
Article number | 8402198 |
Pages (from-to) | 1445-1448 |
Number of pages | 4 |
Journal | IEEE Electron Device Letters |
Volume | 39 |
Issue number | 9 |
DOIs | |
State | Published - Sep 2018 |
Keywords
- FinFET
- Junctionless FET
- ferroelectric FET
- neuromorphic computing system
- synapse