A Compact Model of Non-Volatile Ferroelectric Tunnel Fet with Ambipolarity for in-Memory-Computing Based Edge AI

The novel ferroelectric tunnel FET (FeTFET) with ambipolarity has attracted much attention for area- and energy-efficient edge AI applications. To facilitate the circuit simulation based on the new device, in this work, a compact FeTFET model is established through calculating the ambipolar current based on surface potential and tunneling paths of both the source and drain junctions under all-range of gate bias, and the ferroelectric model based on dynamic Preisach model is also included in the gate stack with non-volatility. Based on the proposed model, the FeTFET demonstrates XNOR operation in one device, showing its high applicability for further FeTFET-based circuit design and simulation.