• DocumentCode
    3087343
  • Title

    Noc Architecture Study with DFG Model

  • Author

    Jia, Liu ; Zheying, Li ; Shuo, Li

  • Author_Institution
    Inst. of Micro Electron. Applic. Technol., Beijing Union Univ., Beijing, China
  • fYear
    2010
  • fDate
    17-19 Sept. 2010
  • Firstpage
    903
  • Lastpage
    906
  • Abstract
    Generic reconfigurable network on chip (GRNoC) is an advanced technology of application specified SoC design for digital signal process system (DSPS). A novel GRNoC mapping method based on data flow graph (DFG) is addressed in this paper. For modules of heterogeneous processors, central memory, and IPs (intellectual properties), DFG model analysis shows that DFG model provides important data transmission properties included the direction and contents of data transmitting, requirements of synchronization and speed of data transmission. The DFG model combined with graph model of GRNoC, therefore, can be the base of route mapping design for the GRNoC. In addition, node architecture of simple router used in GRNoC is also proposed in this paper. The simple router can increases the properties of data transmission in GRNoC and is more suitable for mapping design with DFG model.
  • Keywords
    data flow graphs; digital signal processing chips; network routing; network-on-chip; synchronisation; DSPS; NoC architecture; application specified SoC design; central memory; data flow graph; data transmission properties; data transmission speed; digital signal process system; generic reconfigurable network-on-chip; heterogeneous processors; intellectual properties; node architecture; route mapping design; synchronization; Computer architecture; Data communication; Data models; Digital signal processing; IP networks; Program processors; Tiles; DFG; NoC; SoC;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Pervasive Computing Signal Processing and Applications (PCSPA), 2010 First International Conference on
  • Conference_Location
    Harbin
  • Print_ISBN
    978-1-4244-8043-2
  • Electronic_ISBN
    978-0-7695-4180-8
  • Type

    conf

  • DOI
    10.1109/PCSPA.2010.223
  • Filename
    5635840