Skip to main navigation Skip to search Skip to main content

Dual-port SDRAM architecture for low-power communication of internet-of-things devices

  • Sungchan Kim
  • , Hoeseok Yang*
  • *Corresponding author for this work
    • Ajou University

    Research output: Contribution to journalJournal articlepeer-review

    Abstract

    In this paper, we propose a dual-port SDRAM architecture as an energy-efficient inter-processor communication solution tailored to Internet-of-Things (IoTs) devices. Since a single memory chip plays the role of the local memories and the shared memory for both processors, it gives a simpler solution than a typical architecture that uses an additional dual-port SRAM as shared memory. To minimize the non-negligible synchronization overhead, we also propose two optimization techniques by exploiting the communication patterns of a target application: lock-priority and static-copy. Experiments on a virtual prototyping system show promising results, in which we achieved about 25-50% performance gain as well as about 40% power saving from the proposed optimization techniques compared with the existing architecture.

    Original languageEnglish
    Pages (from-to)893-903
    Number of pages11
    JournalJournal of Semiconductor Technology and Science
    Volume17
    Issue number6
    DOIs
    StatePublished - 2017.12

    UN SDGs

    This output contributes to the following UN Sustainable Development Goals (SDGs)

    1. SDG 7 - Affordable and Clean Energy
      SDG 7 Affordable and Clean Energy
    2. SDG 9 - Industry, Innovation, and Infrastructure
      SDG 9 Industry, Innovation, and Infrastructure

    Keywords

    • Dual-port SDRAM
    • Internet-of-things
    • Low-power communication
    • Memory architecture
    • Memory controller

    Quacquarelli Symonds(QS) Subject Topics

    • Materials Science
    • Engineering - Electrical & Electronic
    • Engineering - Petroleum

    Fingerprint

    Dive into the research topics of 'Dual-port SDRAM architecture for low-power communication of internet-of-things devices'. Together they form a unique fingerprint.

    Cite this