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.

A new method for traveling memory address based on vector function

An Energy Efficient Floating Point Computing Infrastructure Embedding Ferroelectric Field Effect Transistor Based Ternary Content Addressable Memories

CSA-CiM: Enhancing Multi-Functional Computing-in-Memory with Configurable Sense Amplifiers

A Technique to Exploit Memory Locality for Fast Instruction Set Simulation

MBSA: a Lightweight and Flexible Storage Architecture for Virtual Machines

Universal Address Sequence Generator for Memory Built-in Self-test

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

Memory Grouping for the Built-In Self-Test of Three-Dimensional Integrated Circuits

Flatfish: A Reinforcement Learning Approach for Application-Aware Address Mapping

On BTI Aging Rejuvenation in Memory Address Decoders

Efficient Utilization Of Vector Registers To Improve Fft Performance On Simd Microprocessors

Built-In Self-Test Architecture Enabling Diagnosis for Massive Embedded Memory Banks in Large SoCs

CAMA: Energy and Memory Efficient Automata Processing in Content-Addressable Memories

A Power Efficient BIST TPG Method on Don't Care Bit Based 2-D Adjusting and Hamming Distance Based 2-D Reordering

Best Memory Architecture Exploration under Parameters Variations accelerated with Machine Learning

A Power Efficient Architecture with Optimized Parallel Memory Accessing for Feature Generation

Efficient Memory Partitioning For Parallel Data Access Via Data Reuse

An Efficient Transparent Test Scheme for Embedded Word-Oriented Memories

Data Buffer Research Based on Base Address Register Mapping

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