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.

Complementary Memory Cell Based on Field-Programmable Ferroelectric Diode for Ultra-Low Power Current-SA Free BNN Applications

In-Memory-Computing and In-memory Cosine Similarity Search with Memory Diode Featuring Low Latency and High Energy Efficiency

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

Design of Ferroelectric FET-Based Capacitive-Coupling Computing-In-Memory for Binary Neural Networks

HfO 2 -based Junctionless FeFET Array on FDSOI Platform for NAND Memory Application

HfO2-Based Junctionless FeFET Array on FDSOI Platform for NAND Memory Applications

A ferroelectric fin diode for robust non-volatile memory

Two-dimensional ferroelectric channel transistors integrating ultra-fast memory and neural computing

Multilevel Ferroelectric Domain Wall Memory for Neuromorphic Computing

Dual-Ferroelectric-Coupling-Engineered Two-Dimensional Transistors for Multifunctional In-Memory Computing

Enabling Low-Power Charge-Domain Nonvolatile Computing-in-Memory (CIM) With Ferroelectric Memcapacitor

Enabling Lower-Power Charge-Domain Nonvolatile In-Memory Computing with Ferroelectric FETs

Non-Volatile Field-Effect Transistors Enabled by Oxygen Vacancy-Related Dipoles for Memory and Synapse Applications

Organic Nonvolatile 2T Memory Cell Employing a NOT‐Gate‐Like Architecture Toward Binary Output Level With Enhanced Noise Tolerance

A Low-Power Vertical Dual-Gate Neurotransistor with Short-Term Memory for High Energy-Efficient Neuromorphic Computing

Two-terminal Vertical Memory Cell for Cross-Point Static Random Access Memory Applications

Binary Convolutional Neural Network Based on 3d Field-Programmable Ferroelectric Diode Array

Ultralow-power in-memory computing based on ferroelectric memcapacitor network

Multifunctional Computing-in-memory SRAM Cells Based on Two-Surface-channel MoS2 Transistors

In‐Memory Computing of Multilevel Ferroelectric Domain Wall Diodes at LiNbO3 Interfaces

Enhancing Memory Window Efficiency of Ferroelectric Transistor for Neuromorphic Computing via Two‐Dimensional Materials Integration