Notes on the complexity of systolic programs

Stephen Taylor, Lisa Hellerstein, Shmuel Safra, Ehud Shapiro

    Research output: Contribution to journalArticle

    Abstract

    This paper presents two notes which discuss basic theoretical issues concerning the systolic programming methodology and demonstrates simple techniques which can by used to structure communication. The first note shows how the complexity of two simple algorithms is adversely affected by the cost of data movement in a parallel system. Programming techniques which introduce pipelining are used to overcome the problem and improve the complexity. The second note shows an upper bound on the speedup which can be obtained using the methodology on a mesh connected architecture. This bound is stated in terms of the sequential lower bound for the algorithm concerned.

    Original languageEnglish (US)
    Pages (from-to)250-265
    Number of pages16
    JournalJournal of Parallel and Distributed Computing
    Volume4
    Issue number3
    DOIs
    StatePublished - 1987

    Fingerprint

    Programming
    Pipelining
    Methodology
    Parallel Systems
    Speedup
    Mesh
    Lower bound
    Upper bound
    Communication
    Costs
    Demonstrate
    Architecture
    Movement

    ASJC Scopus subject areas

    • Computer Science Applications
    • Hardware and Architecture
    • Control and Systems Engineering

    Cite this

    Notes on the complexity of systolic programs. / Taylor, Stephen; Hellerstein, Lisa; Safra, Shmuel; Shapiro, Ehud.

    In: Journal of Parallel and Distributed Computing, Vol. 4, No. 3, 1987, p. 250-265.

    Research output: Contribution to journalArticle

    Taylor, S, Hellerstein, L, Safra, S & Shapiro, E 1987, 'Notes on the complexity of systolic programs', Journal of Parallel and Distributed Computing, vol. 4, no. 3, pp. 250-265. https://doi.org/10.1016/0743-7315(87)90036-0
    Taylor, Stephen ; Hellerstein, Lisa ; Safra, Shmuel ; Shapiro, Ehud. / Notes on the complexity of systolic programs. In: Journal of Parallel and Distributed Computing. 1987 ; Vol. 4, No. 3. pp. 250-265.
    @article{05f141913b9747e58fcc2ae2d29e1de4,
    title = "Notes on the complexity of systolic programs",
    abstract = "This paper presents two notes which discuss basic theoretical issues concerning the systolic programming methodology and demonstrates simple techniques which can by used to structure communication. The first note shows how the complexity of two simple algorithms is adversely affected by the cost of data movement in a parallel system. Programming techniques which introduce pipelining are used to overcome the problem and improve the complexity. The second note shows an upper bound on the speedup which can be obtained using the methodology on a mesh connected architecture. This bound is stated in terms of the sequential lower bound for the algorithm concerned.",
    author = "Stephen Taylor and Lisa Hellerstein and Shmuel Safra and Ehud Shapiro",
    year = "1987",
    doi = "10.1016/0743-7315(87)90036-0",
    language = "English (US)",
    volume = "4",
    pages = "250--265",
    journal = "Journal of Parallel and Distributed Computing",
    issn = "0743-7315",
    publisher = "Academic Press Inc.",
    number = "3",

    }

    TY - JOUR

    T1 - Notes on the complexity of systolic programs

    AU - Taylor, Stephen

    AU - Hellerstein, Lisa

    AU - Safra, Shmuel

    AU - Shapiro, Ehud

    PY - 1987

    Y1 - 1987

    N2 - This paper presents two notes which discuss basic theoretical issues concerning the systolic programming methodology and demonstrates simple techniques which can by used to structure communication. The first note shows how the complexity of two simple algorithms is adversely affected by the cost of data movement in a parallel system. Programming techniques which introduce pipelining are used to overcome the problem and improve the complexity. The second note shows an upper bound on the speedup which can be obtained using the methodology on a mesh connected architecture. This bound is stated in terms of the sequential lower bound for the algorithm concerned.

    AB - This paper presents two notes which discuss basic theoretical issues concerning the systolic programming methodology and demonstrates simple techniques which can by used to structure communication. The first note shows how the complexity of two simple algorithms is adversely affected by the cost of data movement in a parallel system. Programming techniques which introduce pipelining are used to overcome the problem and improve the complexity. The second note shows an upper bound on the speedup which can be obtained using the methodology on a mesh connected architecture. This bound is stated in terms of the sequential lower bound for the algorithm concerned.

    UR - http://www.scopus.com/inward/record.url?scp=0023366247&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0023366247&partnerID=8YFLogxK

    U2 - 10.1016/0743-7315(87)90036-0

    DO - 10.1016/0743-7315(87)90036-0

    M3 - Article

    VL - 4

    SP - 250

    EP - 265

    JO - Journal of Parallel and Distributed Computing

    JF - Journal of Parallel and Distributed Computing

    SN - 0743-7315

    IS - 3

    ER -