Abstract:The exponential growth of data across various domains of human society necessitates the rapid and efficient data processing. In many contemporary data-intensive applications, similarity computation (SC) is one of the most fundamental and indispensable operations. In recent years, In-memory computing (IMC) architectures have been designed to accelerate SC by reducing data movement costs, however, they encounter challenges with signal domain conversion, variation sensitivity, and limited precision. This paper proposes a ferroelectric FET (FeFET) based time-domain (TD) associative memory (AM) for energy efficient SC. Such TD design can convert its output (i.e., time interval) to digits with relatively simple sensing circuitry thus saves large amount of area and energy compared with conventional IMC designs that process analog voltage/current signals. The variable-capacitance (VC) delay chain structure in our design supports quantitative SC and enhances robustness against variations. Furthermore, by exploiting multi-domain ferroelctric FET (FeFET), our design is capable of performing SC on vectors with multi-bit element, enabling support for higher-precision algorithms. Simulation results show that the proposed TD-AM achieves 13.8x/1.47x energy saving of our design compared to CMOS/NVM based TD-IMC designs. Additionally, our design exhibits good robustness in monte carlo simulation with variation extracted from experimental measurements. Investigation on precision of hyperdimensional computing (HDC) show that higher element precision reduces the size of HDC model when considering to achieve same accuracy, indicating an improved efficiency. Benchmarkings against GPU demonstrate in general 2/3 orders of magnitude speedup/energy efficiency improvement of our design. Our proposed multi-bit TD-AM promises energy-efficient quantitative SC for diverse intensive data processing application, especially in energy-constrained scenarios.

Fully Ferroelectric-FETs Reservoir Computing Network for Temporal and Random Signal Processing

All-ferroelectric implementation of reservoir computing

Switching Dynamics in Anti‐Ferroelectric Transistor for Multimodal Reservoir Computing

Physical reservoirs based on MoS 2 –HZO integrated ferroelectric field-effect transistors for reservoir computing systems

FeFET-Based In-Memory Hyperdimensional Encoding Design

Dynamic Memristor-based Reservoir Computing for High-Efficiency Spatiotemporal Signal Processing

Analysis and Fully Memristor-based Reservoir Computing for Temporal Data Classification

Cost‐Effective and Fully Hardware‐Oriented Reservoir Computing Based on IGZO/HZO Ferroelectric Thin‐Film Transistor with Electrically and Optically Distinguishable States

Thin-film transistor for temporal self-adaptive reservoir computing with closed-loop architecture

Enhancement of Multi-Timescale Reservoir Computing Based on Homogeneous Hardware Platform with Controllable Ionic Dynamics

A FeFET-based Time-Domain Associative Memory for Multi-bit Similarity Computation

An Optoelectronic Reservoir Computing for Temporal Information Processing

Multilayer Reservoir Computing Based on Ferroelectric α-In2 Se3 for Hierarchical Information Processing

Dynamic memristor-based reservoir computing for high-efficiency temporal signal processing

Multilayer Reservoir Computing Based on Ferroelectric Α‐in2se3 for Hierarchical Information Processing

Low Power and Temperature-Resilient Compute-In-Memory Based on Subthreshold-FeFET

Reservoir computing using dynamic memristors for temporal information processing

Energy- and Area-efficient Fe-FinFET-based Time-Domain Mixed-Signal Computing in Memory for Edge Machine Learning

Performance improvement in reservoir computing by using HfZrO2 FeFETs through operating voltage optimization

Analog reservoir computing via ferroelectric mixed phase boundary transistors

Ultralow-Power Compact Artificial Synapse Based on a Ferroelectric Fin Field-Effect Transistor for Spatiotemporal Information Processing