Compute SNR-boosted 22 nm MRAM-based In-memory Computing Macro using Statistical Error Compensation

Saion K. Roy, Han Mo Ou, Mostafa G. Ahmed, Peter Deaville, Bonan Zhang, Naveen Verma, Pavan K. Hanumolu, Naresh R. Shanbhag

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The accuracy of eNVM in-memory computing (IMC) designs is primarily limited by analog non-idealities. This paper presents an MRAM IMC macro in 22nm equipped with offset-compensating current sensing and a low-overhead statistical error compensation (SEC) block to boost its compute signal-to-noise ratio (SNR). The compute SNR reduces with an increase in inner-dimension of the MVM. An SEC-enabled SNR boost of 2.7-to-6 dB is obtained over different operating points. This boost can be traded-off to realize a 5 × decrease in energy/lb-OP including an SEC energy overhead of 0.S%. Finally, we demonstrate an SEC-enabled neural network (NN) accuracy boost from 74.8% to 82.0% for CIFAR-10 over ResNet20 without on-chip training.

Original languageEnglish (US)
Title of host publicationESSCIRC 2023 - IEEE 49th European Solid State Circuits Conference
PublisherIEEE Computer Society
Pages25-28
Number of pages4
ISBN (Electronic)9798350304206
DOIs
StatePublished - 2023
Event49th IEEE European Solid State Circuits Conference, ESSCIRC 2023 - Lisbon, Portugal
Duration: Sep 11 2023Sep 14 2023

Publication series

NameEuropean Solid-State Circuits Conference
Volume2023-September
ISSN (Print)1930-8833

Conference

Conference49th IEEE European Solid State Circuits Conference, ESSCIRC 2023
Country/TerritoryPortugal
CityLisbon
Period9/11/239/14/23

Keywords

  • Compute SNR
  • In-Memory Computing
  • MRAM
  • eNVM

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Compute SNR-boosted 22 nm MRAM-based In-memory Computing Macro using Statistical Error Compensation'. Together they form a unique fingerprint.

Cite this