The Future is Analog: Energy-Efficient Cognitive Network Functions over Memristor-Based Analog Computations

Saad Saleh*, Boris Koldehofe

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

1 Citation (Scopus)

Abstract

Current network functions build heavily on fixed programmed rules and lack capacity to support more expressive learning models, e.g. brain-inspired Cognitive computational models using neuromorphic computations. The major reason for this shortcoming is the huge energy consumption and limitation in expressiveness by the underlying TCAM-based digital packet processors. In this research, we show that recent emerging technologies from the analog domain have a high potential in supporting network functions with energy efficiency and more expressiveness, so called cognitive functions. We propose an analog packet processing architecture building on a novel technology named Memristors. We develop a novel analog match-action memory called Probabilistic Content-Addressable Memory (pCAM) for supporting deterministic and probabilistic match functions. We develop the programming abstractions and show the support of pCAM for an active queue management-based analog network function. The analysis over an experimental dataset of a memristor chip showed only 0.01 fJ/bit/cell of energy consumption for corresponding analog computations which is 50 times less than digital computations.
Original languageEnglish
Title of host publicationProceedings of The 22nd ACM Workshop on Hot Topics in Networks (HotNets’23)
PublisherACM New York, NY, USA
Pages254–262
Number of pages9
ISBN (Print)979-8-4007-0415-4
DOIs
Publication statusPublished - 28-Nov-2023
EventHotNets 2023: Twenty-Second ACM Workshop on Hot Topics in Networks - Cambridge, Massachusetts, United States
Duration: 28-Nov-202329-Nov-2023

Conference

ConferenceHotNets 2023
Country/TerritoryUnited States
City Cambridge, Massachusetts
Period28/11/202329/11/2023

Keywords

  • Network functions
  • Energy efficiency
  • Memristors

Fingerprint

Dive into the research topics of 'The Future is Analog: Energy-Efficient Cognitive Network Functions over Memristor-Based Analog Computations'. Together they form a unique fingerprint.

Cite this