Enhancing the Accuracy of 6T SRAM-based In-Memory Architecture via Maximum Likelihood Detection

Hyungyo Kim, Naresh Shanbhag

Research output: Contribution to journalArticlepeer-review

Abstract

This paper presents a statistical signal processing-based algorithmic approach to enhance the compute signal-to-noise ratio (compute SNR) of 6T SRAM-based analog in-memory computing (IMC) architectures which have recently emerged as an attractive alternative to mainstream digital accelerators for machine learning workloads due to their superior energy efficiency and compute densities. However, today, the compute SNR of analog IMCs is limited by device parameter variations and noise. To overcome this limitation, we propose a maximum likelihood (ML)-based statistical error compensation (MLEC) technique to improve the accuracy of binary dot-products (DPs) realized in 6T (six transistor) SRAM-based analog IMC architectures. The MLEC method involves exploiting the symmetric nature of the 6T SRAM bitcell to extract multiple observations efficiently and employ them for detection purposes. MLEC methods involving two (MLEC-2) and four (MLEC-4) observations are proposed along with efficient architectures to realize them in hardware, e.g., distribution-aware and energy-aware approximations of MLEC-4. Simulations in a commercial 28nm CMOS process demonstrate that the proposed methods increase the compute SNR for the commonly used 144-dimensional DP by 5dB-to-12dB. This improvement in the bank-level compute SNR leads to a network-level accuracy improvement of up to 11% when a ResNet-20 (CIFAR-10 dataset) network is implemented on the IMC. Employing energy models of the IMC, the energy overhead of MLEC is estimated to lie between 3%-to-10% resulting in up to 45.6% and 18% increase in energy efficiency (1b-TOPS/W) for a target SNR of 20dB and ResNet-20 accuracy of 90% on the CIFAR-10 dataset, respectively, compared to a conventional (uncompensated) 6T SRAM-based IMC.

Original languageEnglish (US)
Pages (from-to)1-13
Number of pages13
JournalIEEE Transactions on Signal Processing
DOIs
StateAccepted/In press - 2024
Externally publishedYes

Keywords

  • Computational modeling
  • Computer architecture
  • in-memory computing
  • maximum likelihood detection
  • Signal to noise ratio
  • Standards
  • statistical error compensation
  • Task analysis
  • Transistors
  • Vectors

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Enhancing the Accuracy of 6T SRAM-based In-Memory Architecture via Maximum Likelihood Detection'. Together they form a unique fingerprint.

Cite this