Skip to main content
Video s3
    Details
    Presenter(s)
    Anil Korkmaz Headshot
    Display Name
    Anil Korkmaz
    Affiliation
    Affiliation
    Texas A&M University
    Country
    Country
    Türkiye
    Author(s)
    Display Name
    Gianluca Zoppo
    Affiliation
    Affiliation
    Politecnico di Torino
    Display Name
    Anil Korkmaz
    Affiliation
    Affiliation
    Texas A&M University
    Display Name
    Francesco Marrone
    Affiliation
    Affiliation
    Politecnico di Torino
    Display Name
    Samuel Palermo
    Affiliation
    Affiliation
    Texas A&M University
    Display Name
    Fernando Corinto
    Affiliation
    Affiliation
    Politecnico di Torino
    Display Name
    Richard Williams
    Affiliation
    Affiliation
    Texas A&M University
    Abstract

    Recently, several research groups have demonstrated that MVMs and linear systems can be performed analytically in a single time step with memristor crossbars. These two approaches can both be applied to Markov chains. We present circuit models for open-loop and feedback configurations, and perform analyses that include memristor programming errors, thermal noise sources and element nonidealities in realistic circuit simulations to determine both the precision and accuracy of the analog solutions. We provide mathematical tools to describe the trade-offs in the circuit model between power consumption and magnitude of errors. We compare the two approaches by analyzing Markov chains applications.

    Slides
    • Analog Solutions of Discrete Markov Chains via Memristor Crossbars (application/pdf)