• DocumentCode
    3126518
  • Title

    A new approach for clustering network modules for delay minimization

  • Author

    Karayiannis, D. ; Tragoudas, S.

  • Author_Institution
    Dept. of Comput. Sci., Southern Illinois Univ., Carbondale, IL, USA
  • fYear
    1995
  • fDate
    1995
  • Firstpage
    325
  • Lastpage
    332
  • Abstract
    In recent years there has been an extensive interest in clustering the modules of a network so that the maximum delay from any primary input to any primary output is minimized. Clusters have a maximum capacity, pin limitations and modules have different implementations. Furthermore, different implementations may have different areas. All existing CAD frameworks initially select an implementation of each module, and at a later stage they cluster the modules. We present an approach that clusters the nodes, while considering their alternative implementations, so that we further minimize the maximum delay after the clustering. Our approach is based on optimal algorithms for restricted versions of this complex problem in circuit design, and outperforms the conventional approach, which first obtains an implementation for each circuit module without considering clustering and then, in a later step, performs clustering.
  • Keywords
    circuit CAD; circuit optimisation; combinational circuits; delays; network synthesis; CAD; circuit design; combinational circuit; delay minimization; maximum capacity; maximum delay; network module clustering; optimal algorithms; pin limitations; Circuit synthesis; Clustering algorithms; Computer science; Delay; Electronics packaging; Integrated circuit interconnections; Law; Minimization; Partitioning algorithms; Pins;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Systems Engineering of Computer Based Systems, 1995., Proceedings of the 1995 International Symposium and Workshop on
  • Conference_Location
    Tucson, AZ, USA
  • Print_ISBN
    0-7803-2531-1
  • Type

    conf

  • DOI
    10.1109/ECBS.1995.521872
  • Filename
    521872