Near-memory computing: past, present, and future

Gagandeep Singh (Corresponding author), Lorenzo Chelini, Stefano Corda, Ahsan Javed Awan, Sander Stuijk, Roel Jordans, Henk Corporaal, Albert-Jan Boonstra

Onderzoeksoutput: Bijdrage aan tijdschriftTijdschriftartikelAcademicpeer review

Uittreksel

The conventional approach of moving data to the CPU for computation has become a significant performance bottleneck for emerging scale-out data-intensive applications due to their limited data reuse. At the same time, the advancement in 3D integration technologies has made the decade-old concept of coupling compute units close to the memory — called near-memory computing (NMC) — more viable. Processing right at the “home” of data can significantly diminish the data movement problem of data-intensive applications. In this paper, we survey the prior art on NMC across various dimensions (architecture, applications, tools, etc.) and identify the key challenges and open issues with future research directions. We also provide a glimpse of our approach to near-memory computing that includes i) NMC specific microarchitecture independent application characterization ii) a compiler framework to offload the NMC kernels on our target NMC platform and iii) an analytical model to evaluate the potential of NMC.

TaalEngels
Artikelnummer102868
Aantal pagina's15
TijdschriftMicroprocessors and Microsystems
Volume71
DOI's
StatusGepubliceerd - 1 nov 2019

Vingerafdruk

Data storage equipment
Program processors
Analytical models
Processing

Trefwoorden

    Citeer dit

    @article{13c1c5ccd301431fb2a3bbe5d6a3a03f,
    title = "Near-memory computing: past, present, and future",
    abstract = "The conventional approach of moving data to the CPU for computation has become a significant performance bottleneck for emerging scale-out data-intensive applications due to their limited data reuse. At the same time, the advancement in 3D integration technologies has made the decade-old concept of coupling compute units close to the memory — called near-memory computing (NMC) — more viable. Processing right at the “home” of data can significantly diminish the data movement problem of data-intensive applications. In this paper, we survey the prior art on NMC across various dimensions (architecture, applications, tools, etc.) and identify the key challenges and open issues with future research directions. We also provide a glimpse of our approach to near-memory computing that includes i) NMC specific microarchitecture independent application characterization ii) a compiler framework to offload the NMC kernels on our target NMC platform and iii) an analytical model to evaluate the potential of NMC.",
    keywords = "Application characterization, Computer architecture, Data-centric computing, Modeling, Near-memory computing, Survey",
    author = "Gagandeep Singh and Lorenzo Chelini and Stefano Corda and Awan, {Ahsan Javed} and Sander Stuijk and Roel Jordans and Henk Corporaal and Albert-Jan Boonstra",
    year = "2019",
    month = "11",
    day = "1",
    doi = "10.1016/j.micpro.2019.102868",
    language = "English",
    volume = "71",
    journal = "Microprocessors and Microsystems",
    issn = "0141-9331",
    publisher = "Elsevier",

    }

    Near-memory computing : past, present, and future. / Singh, Gagandeep (Corresponding author); Chelini, Lorenzo; Corda, Stefano; Awan, Ahsan Javed; Stuijk, Sander; Jordans, Roel; Corporaal, Henk; Boonstra, Albert-Jan.

    In: Microprocessors and Microsystems, Vol. 71, 102868, 01.11.2019.

    Onderzoeksoutput: Bijdrage aan tijdschriftTijdschriftartikelAcademicpeer review

    TY - JOUR

    T1 - Near-memory computing

    T2 - Microprocessors and Microsystems

    AU - Singh,Gagandeep

    AU - Chelini,Lorenzo

    AU - Corda,Stefano

    AU - Awan,Ahsan Javed

    AU - Stuijk,Sander

    AU - Jordans,Roel

    AU - Corporaal,Henk

    AU - Boonstra,Albert-Jan

    PY - 2019/11/1

    Y1 - 2019/11/1

    N2 - The conventional approach of moving data to the CPU for computation has become a significant performance bottleneck for emerging scale-out data-intensive applications due to their limited data reuse. At the same time, the advancement in 3D integration technologies has made the decade-old concept of coupling compute units close to the memory — called near-memory computing (NMC) — more viable. Processing right at the “home” of data can significantly diminish the data movement problem of data-intensive applications. In this paper, we survey the prior art on NMC across various dimensions (architecture, applications, tools, etc.) and identify the key challenges and open issues with future research directions. We also provide a glimpse of our approach to near-memory computing that includes i) NMC specific microarchitecture independent application characterization ii) a compiler framework to offload the NMC kernels on our target NMC platform and iii) an analytical model to evaluate the potential of NMC.

    AB - The conventional approach of moving data to the CPU for computation has become a significant performance bottleneck for emerging scale-out data-intensive applications due to their limited data reuse. At the same time, the advancement in 3D integration technologies has made the decade-old concept of coupling compute units close to the memory — called near-memory computing (NMC) — more viable. Processing right at the “home” of data can significantly diminish the data movement problem of data-intensive applications. In this paper, we survey the prior art on NMC across various dimensions (architecture, applications, tools, etc.) and identify the key challenges and open issues with future research directions. We also provide a glimpse of our approach to near-memory computing that includes i) NMC specific microarchitecture independent application characterization ii) a compiler framework to offload the NMC kernels on our target NMC platform and iii) an analytical model to evaluate the potential of NMC.

    KW - Application characterization

    KW - Computer architecture

    KW - Data-centric computing

    KW - Modeling

    KW - Near-memory computing

    KW - Survey

    UR - http://www.scopus.com/inward/record.url?scp=85070573731&partnerID=8YFLogxK

    U2 - 10.1016/j.micpro.2019.102868

    DO - 10.1016/j.micpro.2019.102868

    M3 - Article

    VL - 71

    JO - Microprocessors and Microsystems

    JF - Microprocessors and Microsystems

    SN - 0141-9331

    M1 - 102868

    ER -