• Home
  • Search
  • Neural-PIM: Efficient Processing-In-Memory with Neural Approximation of Peripherals
  • Cite Icon20
  • https://doi.org/10.1109/tc.2021.3122905Copy DOI Icon

Neural-PIM: Efficient Processing-In-Memory with Neural Approximation of Peripherals

Show More
  • Abstract
  • Literature Map
  • References
  • Citations
  • Similar Papers
Abstract

Processing-in-memory (PIM) architectures have demonstrated great potential in accelerating numerous deep learning tasks. Particularly, resistive random-access memory (RRAM) devices provide a promising hardware substrate to build PIM accelerators due to their abilities to realize efficient in-situ vector-matrix multiplications (VMMs). However, existing PIM accelerators suffer from frequent and energy-intensive analog-to-digital (A/D) conversions, severely limiting their performance. This paper presents a new PIM architecture to efficiently accelerate deep learning tasks by minimizing the required A/D conversions with analog accumulation and neural approximated peripheral circuits. We first characterize the different dataflows employed by existing PIM accelerators, based on which a new dataflow is proposed to remarkably reduce the required A/D conversions for VMMs by extending shift and add (S+A) operations into the analog domain before the final quantizations. We then leverage a neural approximation method to design both analog accumulation circuits (S+A) and quantization circuits (ADCs) with RRAM crossbar arrays in a highly-efficient manner. Finally, we apply them to build an RRAM-based PIM accelerator (i.e., \textbf{Neural-PIM}) upon the proposed analog dataflow and evaluate its system-level performance. Evaluations on different benchmarks demonstrate that Neural-PIM can improve energy efficiency by 5.36x (1.73x) and speed up throughput by 3.43x (1.59x) without losing accuracy, compared to the state-of-the-art RRAM-based PIM accelerators, i.e., ISAAC (CASCADE).

Similar Papers
  • Research Article
  • Citations6

Improving the operational characteristic stability in Al/Au/ZnO/Al resistive random access memory devices

  • Feb 24, 2016
  • Japanese Journal of Applied Physics
  • Cheng-Yen Wu +3
  • Research Article
  • Citations3

Resistive Switching Characteristics of 10-nm-Thick Amorphous HoScO x Films Doped with Nb and Zn

  • Dec 22, 2016
  • Journal of Electronic Materials
  • Sea-Fue Wang +4
  • Research Article
  • Citations2

Modulation of the electrical characteristics on poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate)-based resistive random access memory device by the impact of top electrode materials

  • Mar 05, 2022
  • Thin Solid Films
  • Mehr Khalid Rahmani +4
  • Research Article
  • Citations27

Switching characteristics of ZnO based transparent resistive random access memory devices grown by pulsed laser deposition

  • May 21, 2010
  • physica status solidi c
  • Pankaj Misra +2
  • Research Article

Towards a Memristor‐Based Circuit Implementation of the Hindmarsh–Rose Model

  • Sep 08, 2025
  • International Journal of Circuit Theory and Applications
  • Sebastian Jenderny +3
  • Research Article
  • Citations12

Effects of Film Orientation on Power Consumption, Thermal Stability, and Reliability of Aluminum Nitride Resistive Random Access Memory Devices

  • Nov 01, 2019
  • IEEE Transactions on Electron Devices
  • Chun-Cheng Lin +5
  • Research Article
  • Citations8

Factors that control stability, variability, and reliability issues of endurance cycle in ReRAM devices: A phase field study

  • May 10, 2022
  • Journal of Applied Physics
  • Arijit Roy +2
  • Research Article
  • Citations32

Resistive random access memory and its applications in storage and nonvolatile logic**The resistive random access memory (RRAM) device has been widely studied due to its excellent memory characteristics and great application potential in different fields. In this paper, resistive switching materials, switching mechanism, and memory characteristics of RRAM are discussed. Recent research progress of RRAM in high-density storage and nonvolatile

  • Jul 01, 2017
  • Journal of Semiconductors
  • Dongbin Zhu +5
  • Research Article
  • Citations3

Effects of Ambient and Annealing Temperature in HfO2 Based RRAM Device Modeling and Circuit-Level Implementation

  • Feb 01, 2022
  • ECS Journal of Solid State Science and Technology
  • A Napolean +6
  • Research Article

(Invited) Understanding the Role of Dopants in Transition Metal Oxide Dielectrics for Digital and Analog Resistive Switching

  • May 03, 2013
  • Electrochemical Society Transactions
  • Branden Long +5
  • Research Article
  • Citations19

Low-power multilevel resistive switching in β-Ga2O3 based RRAM devices

  • Dec 02, 2022
  • Nanotechnology
  • Ravi Teja Velpula +2
  • PDF
  • Research Article
  • Citations59

Light-activated resistance switching in SiOx RRAM devices

  • Dec 04, 2017
  • Applied Physics Letters
  • A Mehonic +2
  • Research Article
  • Citations34

Design of wafer-scale uniform Au nanotip array by ion irradiation for enhanced single conductive filament resistive switching

  • Oct 22, 2019
  • Nano Energy
  • Wenqing Li +8
  • Research Article
  • Citations41

Resistive switching mechanism of Ag/ZrO2:Cu/Pt memory cell

  • Jan 26, 2011
  • Applied Physics A
  • Shibing Long +10
  • Research Article
  • Citations13

Stable switching behavior of low-temperature ZrO2 RRAM devices realized by combustion synthesis-assisted photopatterning

  • Jan 17, 2024
  • Journal of Materials Science & Technology
  • Bongho Jang +4
Cactus Communications logo

Copyright 2026 Cactus Communications. All rights reserved.