Improved Pattern Recognition Accuracy of Hardware Neural Network: Deactivating Short Failed Synapse Device by Adopting Ovonic Threshold Switching (OTS)-Based Fuse Device

In analog crosspoint array-based architectures for neural network computation, short failure of a crosspoint synaptic device can cause a serious issue, preventing normal operation of the devices sharing the row and column as well as the peripheral circuits. In this paper, we characterize a 32 × 32 array of PCMO (Pr
_x
Ca
_1-x
MnO
₃
)based resistive random-access memory (RRAM) devices, and obtained short failure statistics. Our analysis shows that inference accuracy of a neural network system can be significantly degraded when only a small number of devices become short-failed in the array. We analyze that the main cause of the pattern recognition accuracy degradation is resulted from the unexpected neural excitation by high current through the shorted device. To prevent such accuracy degradation, we fabricate an ovonic threshold switching (OTS)-based electrical fuse device, and experimentally verify its fuse functionality utilizing the void formation phenomenon in chalcogenide materials under high current condition. After disconnecting short-failed devices with the fuse operation, the recognition accuracy is recovered from 10% to 97%, which is on par with the performance with no short-failed devices in the array.