Ferroelectric Field-Effect Transistors in Analog Applications: Demonstrated in Time-Domain Neuromorphic Computing Cells

In response to diminishing improvements from transistor scaling, the semiconductor field has shifted its focus to architectural innovation and emerging device technologies. Ferroelectric field-effect transistors (FeFETs) are non-volatile memory elements which offer new opportunities for energy-efficient computing, particularly in compute-in-memory (CIM) architectures. This thesis presents a novel FeFET-based computing cell integrated into a time-domain (TD) computing architecture. The proposed design is evaluated using a device model derived from experimental data and compared with competing non-volatile memory implementations. Results demonstrate improvements in area efficiency and performance, highlighting the potential of the FeFET cell for next-generation TD-CIM systems.