Skip to main navigation Skip to search Skip to main content

A Novel Hybrid Transactional Memory Based on Abort Prediction and Adaptive Retry Policy

    Research output: Contribution to conferenceConference paperpeer-review

    Abstract

    This paper proposes a novel hybrid transactional memory scheme based on both abort prediction and an adaptive retry policy. First, the proposed scheme can predict not only conflicts between transactions running concurrently, but also the capacity and other aborts of transactions by collecting the information of previously executed transactions. Second, the proposed scheme can provide an adaptive retry policy based on machine learning algorithms, according to the characteristic of a given workload. Finally, through our experimental performance analysis using STAMP, the proposed scheme shows about 20% better performance than Hybrid NOrec, a hybrid version of the efficient NOrec STM.

    Original languageEnglish
    Title of host publicationProceedings - 2017 IEEE International Conference on Cluster Computing, CLUSTER 2017
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages613-614
    Number of pages2
    ISBN (Electronic)9781538623268
    DOIs
    StatePublished - 2017.09.22
    Event2017 IEEE International Conference on Cluster Computing, CLUSTER 2017 - Honolulu, United States
    Duration: 2017.09.52017.09.8

    Publication series

    NameProceedings - IEEE International Conference on Cluster Computing, ICCC
    Volume2017-September
    ISSN (Print)1552-5244

    Conference

    Conference2017 IEEE International Conference on Cluster Computing, CLUSTER 2017
    Country/TerritoryUnited States
    CityHonolulu
    Period17.09.517.09.8

    Keywords

    • Concurrency control
    • HTM
    • Multi-core in-memory database
    • STM
    • Transactional memory

    Quacquarelli Symonds(QS) Subject Topics

    • Computer Science & Information Systems
    • Data Science

    Fingerprint

    Dive into the research topics of 'A Novel Hybrid Transactional Memory Based on Abort Prediction and Adaptive Retry Policy'. Together they form a unique fingerprint.

    Cite this