Compiler assisted dynamic allocation of finite hardware acceleration resources for parallel tasks

Jari-Matti Mäkelä, Martti Forsell, Ville Leppänen

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

    Abstract

    With the modern chip design facing the so called frequency, power and other walls, multi-core systems have become dominant. Due to the relatively large operating cost of fine-grained parallelism, the systems are often task-oriented. With the recent appearance of more synchronous platforms for parallel computing, we propose a language supported task system optimization for strictly synchronous multi-core architectures equipped with hardware accelerated multi-prefix operations and other finite resources (with respect to concurrent access). The purpose of this system is to dynamically manage the error prone allocation of such resources, yet provide reasonable performance speedups without placing the burden of resource management to the programmer. Our system compiles several specialized versions of accelerated functions that are dynamically picked by the runtime task system based on the tasks' requirements and the resource availability.
    Original languageEnglish
    Title of host publicationCompSysTech '16 Proceedings of the 17th International Conference on Computer Systems and Technologies 2016
    PublisherAssociation for Computing Machinery ACM
    Pages49-56
    ISBN (Print)978-1-4503-4182-0
    DOIs
    Publication statusPublished - 2016
    MoE publication typeA4 Article in a conference publication
    Event17th International Conference on Computer Systems and Technologies, CompSysTech 2016 - Palermo, Italy
    Duration: 23 Jun 201624 Jun 2016
    Conference number: 17

    Conference

    Conference17th International Conference on Computer Systems and Technologies, CompSysTech 2016
    Abbreviated titleCompSysTech 2016
    CountryItaly
    CityPalermo
    Period23/06/1624/06/16

    Keywords

    • active memory
    • flow analysis
    • parallel programming
    • task parallel

    Fingerprint Dive into the research topics of 'Compiler assisted dynamic allocation of finite hardware acceleration resources for parallel tasks'. Together they form a unique fingerprint.

  • Cite this

    Mäkelä, J-M., Forsell, M., & Leppänen, V. (2016). Compiler assisted dynamic allocation of finite hardware acceleration resources for parallel tasks. In CompSysTech '16 Proceedings of the 17th International Conference on Computer Systems and Technologies 2016 (pp. 49-56). Association for Computing Machinery ACM. https://doi.org/10.1145/2983468.2983494