Analog Content-Addressable Memory from Complementary FeFETs

TL;DR

This paper introduces an analog CAM cell using complementary FeFETs that enables high-density, fast, and accurate similarity search, significantly advancing compute-in-memory capabilities for AI and data-intensive tasks.

Contribution

It presents a novel analog CAM design based on FeFETs that performs parallel similarity search with over 40 match windows, improving density and speed over existing architectures.

Findings

Outperforms ternary CAM in few-shot learning accuracy by 5%.

Achieves 3x higher memory density and over 100x faster search speed than CPU/GPU.

Enables 1-step inference in kernel regression, 1000x faster than traditional methods.

Abstract

To address the increasing computational demands of artificial intelligence (AI) and big data, compute-in-memory (CIM) integrates memory and processing units into the same physical location, reducing the time and energy overhead of the system. Despite advancements in non-volatile memory (NVM) for matrix multiplication, other critical data-intensive operations, like parallel search, have been overlooked. Current parallel search architectures, namely content-addressable memory (CAM), often use binary, which restricts density and functionality. We present an analog CAM (ACAM) cell, built on two complementary ferroelectric field-effect transistors (FeFETs), that performs parallel search in the analog domain with over 40 distinct match windows. We then deploy it to calculate similarity between vectors, a building block in the following two machine learning problems. ACAM outperforms ternary CAM (TCAM) when applied to similarity search for few-shot learning on the Omniglot dataset, yielding projected simulation results with improved inference accuracy by 5%, 3x denser memory architecture, and more than 100x faster speed compared to central processing unit (CPU) and graphics processing unit (GPU) per similarity search on scaled CMOS nodes. We also demonstrate 1-step inference on a kernel regression model by combining non-linear kernel computation and matrix multiplication in ACAM, with simulation estimates indicating 1,000x faster inference than CPU and GPU.