• DocumentCode
    992887
  • Title

    Arithmetic unit design for neural accelerators: cost performance issues

  • Author

    Sammut, K.M. ; Jones, S.R.

  • Author_Institution
    Dept. of Electron. & Electr. Eng., Loughborough Univ. of Technol., UK
  • Volume
    44
  • Issue
    10
  • fYear
    1995
  • fDate
    10/1/1995 12:00:00 AM
  • Firstpage
    1256
  • Lastpage
    1260
  • Abstract
    Arithmetic unit design is a key issue when supporting the computational requirements of neural networks. However, there is little quantitative evidence from the study of existing neural accelerators to help choose between arithmetic constructs. This paper presents an assessment of the cost-performance trade-offs between arithmetic constructs for linear neural accelerators
  • Keywords
    digital arithmetic; neural nets; arithmetic unit design; cost performance issues; cost-performance trade-offs; neural accelerators; Arithmetic; Circuits and systems; Clocks; Computer architecture; Computer networks; Costs; Linear accelerators; Logic testing; Neural networks; System testing;
  • fLanguage
    English
  • Journal_Title
    Computers, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    0018-9340
  • Type

    jour

  • DOI
    10.1109/12.467702
  • Filename
    467702