• DocumentCode
    146219
  • Title

    On wiring delays reduction of tree-based FPGA using 3-D fabric

  • Author

    Pangracious, Vinod ; Marrakchi, Mohamed Sahbi ; Mehrez, H. ; Marrakchi, Z.

  • Author_Institution
    LIP6, Univ. Pierre & Marie Curie Paris VI, Paris, France
  • fYear
    2014
  • fDate
    2-5 Sept. 2014
  • Firstpage
    64
  • Lastpage
    69
  • Abstract
    We describe the design and exploration methodology to optimize 3-dimensional (3D) heterogeneous interconnect fabric of Tree-based FPGA (HT-FPGA) by introducing a break-point at a particular tree level interconnect to optimize the speed, power consumption and area. The ability of the flow to decide a horizontal or vertical partitioning of the multilevel programmable tree network based on design specifications is a defining feature. The break-point of vertically partitioned tree is designed to balance the placement of logic blocks and switch blocks into multiple tiers while the break-point of horizontally partitioned tree is designed to optimize the interconnect delay of the programmable tree network. We finally evaluate the performance, area and power consumption of the proposed 3D HT-FPGA using the newly developed flow and show that vertical and horizontally partitioned 3D stacked HT-FPGA improves speed by 16% and 55% respectively. Silicon footprint reduced by 50% for vertical and 46% for horizontal partitioning method and power consumption reduced by 35% compared to 2D counterpart.
  • Keywords
    field programmable gate arrays; integrated circuit design; integrated circuit interconnections; three-dimensional integrated circuits; trees (mathematics); 3-dimensional heterogeneous interconnect fabric; 3D heterogeneous interconnect fabric; 3D stacked HT-FPGA; horizontal partitioning; horizontally partitioned tree; interconnect delay; logic blocks; multilevel programmable tree network; silicon footprint; switch blocks; tree level interconnect; tree-based FPGA; vertical partitioning; vertically partitioned tree; Adders; Delays; Field programmable gate arrays; Optimization; Power demand; Three-dimensional displays; Through-silicon vias;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    System-on-Chip Conference (SOCC), 2014 27th IEEE International
  • Conference_Location
    Las Vegas, NV
  • Type

    conf

  • DOI
    10.1109/SOCC.2014.6948901
  • Filename
    6948901