Co-Design of Algorithms & Hardware for DNNs - Vivienne Sze - LPIRC 2018

  • 87 views


  • Download

    Download

    IEEE.tv is made possible by the Members of IEEE.

    This feature is accessible to IEEE Members only, with an IEEE Account. If you are an IEEE Member please sign in to enable this feature.

    In addition to exclusive access to IEEE.tv programming, IEEE members have file download, and can save favorite videos with myTV.

    Discover all the benefits of IEEE membership!

    Belong to the world's largest technical proffecional society

  • Share
    Conference Highlights
  • #IEEE
  • #LPIRC
  • #2018
  • #low power
  • #image recognition
  • #competition
  • #MLT
  • #MIT
  • #design
  • #algorithm
  • #hardware
  • #deep neural networks
  • #DNN
  • #deep learning
  • #direct metrics
  • #flexible dataflow
  • #NoC

  • Transcript

    IEEE.tv is made possible by the Members of IEEE.

    This feature is accessible to IEEE Members only, with an IEEE Account. If you are an IEEE Member please sign in to enable this feature.

    In addition to exclusive access to IEEE.tv programming, IEEE members have file download, and can save favorite videos with myTV.

    Discover all the benefits of IEEE membership!

    Belong to the world's largest technical proffecional society

Vivienne Sze, Microsystems Technology Laboratories at MIT, takes a look at efficiency optimization of algorithm design and considering energy evaluation when building Deep Neural Networks. Sze presents comparisons of adaptations and various examples of existing DNN architecture on different hardware.

More

  • Published on
  • September 12, 2019

NEXT UP

Avatar

Biography

Videos with