Synaptic SONOS nonvolatile analog memories for artifical neural networks

In this paper, a new synapse cell with nonvolatile SONOS semiconductor memory device is proposed. The memory value, synaptic weights, can be changed incrementally. A novel SONOS synapse is used to read out the stored value. For the purpose of synapse implementation using SONOS NVSM, this work has investigated weight updating characteristics including multiplying characteristics. It is concluded that SONOS synapse cell has excellent characteristics for use as a synapse in artificial neural networks.