• DocumentCode
    494472
  • Title

    Using adaptive routing to compensate for performance heterogeneity

  • Author

    Markovsky, Yury ; Patel, Yatish ; Wawrzynek, John

  • Author_Institution
    Univ. of California, Berkeley, CA
  • fYear
    2009
  • fDate
    10-13 May 2009
  • Firstpage
    12
  • Lastpage
    21
  • Abstract
    Scalable and power efficient multi-core architectures must be performance heterogeneous to accommodate semi-conductor parametric variations and non-uniform access to shared resources. Due to its rate matching, a NoC on a Voltage-Frequency Island architecture can connect cores without forcing each one to give up its own operating point for the chip-wide common worst case. With run-time adaptive routing and task-to-core mapping, a NoC can run at the average not the worst case network saturation bandwidth. These run-time processes compensate for variations because they match application resource requirements with heterogeneous cores and routers. We focus on adaptive routing that simultaneously combats communication load imbalance from on-die variations and application topology. We show that even with static, fixed task-to-core mapping on multi-core architectures affected by stochastic variations, our MATC router increases the expected saturation bandwidth by 7-25% vs Dimension Order router. With systematic variations, the improvements are 5-50%. These gains compensate for saturation bandwidth degradation due to manufacturing variations and help to reduce design guard-bands.
  • Keywords
    network-on-chip; stochastic processes; adaptive routing; communication load imbalance; dimension order router; fixed task-to-core mapping; guard-bands; multicore architectures; on-die variations; saturation bandwidth; stochastic variations; Adaptive systems; Bandwidth; Circuits; Network topology; Network-on-a-chip; Routing; Runtime; Stochastic processes; Telecommunication traffic; Voltage;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Networks-on-Chip, 2009. NoCS 2009. 3rd ACM/IEEE International Symposium on
  • Conference_Location
    San Diego, CA
  • Print_ISBN
    978-1-4244-4142-6
  • Electronic_ISBN
    978-1-4244-4143-3
  • Type

    conf

  • DOI
    10.1109/NOCS.2009.5071440
  • Filename
    5071440