Abstract:Artificial intelligence has made remarkable advancements in recent years, leading to the development of algorithms and models capable of handling ever-increasing amounts of data. The computational demands of these algorithms necessitate circuit and architecture designs that go beyond the von-Neumann paradigm. Content addressable memories (CAMs), which implement parallel associative search functionality within memory blocks to overcome the memory wall bottleneck, have proven to be effective for data-intensive tasks. While current CAM designs have achieved higher storage density and energy efficiency than their CMOS-based counterparts by leveraging emerging non-volatile memories (NVM), most of these implementations are limited to binary storage cells. In this work, we propose SEE-MCAM, scalable and compact multi-bit CAM (MCAM) designs that utilize the three-terminal ferroelectric FET (FeFET) as the proxy. By exploiting the multi-level-cell characteristics of FeFETs, our proposed SEE-MCAM designs enable multi-bit associative search functions and achieve better energy efficiency and performance than existing FeFET-based CAM designs. We validated the functionality of our proposed designs by achieving 3 bits per cell CAM functionality, resulting in 3x improvement in storage density. The area per bit of the proposed SEE-MCAM cell is 8% of the conventional CMOS CAM. We thoroughly investigated the scalability and robustness of the proposed design. Evaluation results suggest that the proposed 2FeFET-1T SEE-MCAM achieves 9.8x more energy efficiency and 1.6x less search latency compared to the CMOS CAM, respectively. When compared to existing MCAM designs, the proposed SEE-MCAM can achieve 8.7x and 4.9x more energy efficiency than ReRAM-based and FeFET-based MCAMs, respectively. Benchmarking results show that our approach provides up to 3 orders of magnitude improvement in speedup and energy efficiency over a GPU implementation in accelerating a novel quantized hyperdimensional computing (HDC) application.

Flash-Based Content Addressable Memory with L2 Distance for Memory-Augmented Neural Network

A Scalable 2T-1Fefet Based Content Addressable Memory Design for Energy Efficient Data Search

A Low-Power In-Memory Multiplication and Accumulation Array with Modified Radix-4 Input and Canonical Signed Digit Weights

Analog Content Addressable Memory using Ferroelectric: A Case Study of Search-in-Memory

Design of Ultracompact Content Addressable Memory Exploiting 1T-1MTJ Cell

Energy Efficient Data Search Design and Optimization Based on a Compact Ferroelectric FET Content Addressable Memory

A Scalable Design of Multi-Bit Ferroelectric Content Addressable Memory for Data-Centric Computing

First Demonstration of Charge-Domain Content Addressable Memory Based on Ferroelectric Capacitive Memory for Reliable and Low-Power One-Shot Learning

SEE-MCAM: Scalable Multi-bit FeFET Content Addressable Memories for Energy Efficient Associative Search

Ferroelectric Ternary Content Addressable Memories for Energy-Efficient Associative Search.

An Ultracompact Single-Ferroelectric Field-Effect Transistor Binary and Multibit Associative Search Engine

Ferroelectric Ternary Content-Addressable Memory for One-Shot Learning

A Novel Small-Signal Ferroelectric Capacitance-Based Content Addressable Memory for Area- and Energy-Efficient Lifelong Learning

Analog Content-Addressable Memory from Complementary FeFETs

Flash Memory Array for Efficient Implementation of Deep Neural Networks

Ferroelectric FET Based In-Memory Computing for Few-Shot Learning.

An Energy-Efficient Architecture for Accelerating Inference of Memory-Augmented Neural Networks

Efficient and Reliable Vector Similarity Search Using Asymmetric Encoding with NAND-Flash for Many-Class Few-Shot Learning

Single-Ferroelectric FET based Associative Memory for Data-Intensive Pattern Matching.

Optimized operation scheme of flash-memory-based neural network online training with ultra-high endurance

Gain Cell-Based Analog Content Addressable Memory for Dynamic Associative tasks in AI