Advanced search
1 file | 591.57 KB Add to list

Dynamic data folding with parameterizable FPGA configurations

Karel Bruneel (UGent) , Wim Heirman (UGent) and Dirk Stroobandt (UGent)
Author
Organization
Abstract
In many applications, subsequent data manipulations differ only in a small set of parameter values. Because of their reconfigurability, FPGAs (field programmable gate arrays) can be configured with a specialized circuit each time the parameter values change. This technique is called dynamic data folding. The specialized circuits are smaller and faster than their generic counterparts. However; the overhead involved in generating the configurations for the specialized circuits at runtime is very large when conventional tools are used, and this overhead will in many cases negate the benefit of using optimized configurations. This article introduces an automatic method for generating runtime parameterizable configurations from arbitrary Boolean circuits. These configurations, in which some of the configuration bits are expressed as a closed-form Boolean expression of a set of parameters, enable very fast run-time specialization, since specialization only involves evaluating these expressions. Our approach is validated on a ternary content-addressable memory (TCAM). We show that the specialized configurations, produced by our method use 2.82 times fewer LUTs than the generic configuration, and even 1.41 times fewer LUTs than the implementation generated by Xilinx Coregen. Moreover, while Coregen needs hand-crafted generators for each type of circuit, our toolflow can be applied to any VHDL design. Using our automatic and generally applicable method, run-time hardware optimization suddenly becomes feasible for a large class of applications.
Keywords
dynamic data folding, Automatic hardware synthesis, FPGA, run-time reconfiguration

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 591.57 KB

Citation

Please use this url to cite or link to this publication:

MLA
Bruneel, Karel, Wim Heirman, and Dirk Stroobandt. “Dynamic Data Folding with Parameterizable FPGA Configurations.” ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS 16.4 (2011): n. pag. Print.
APA
Bruneel, K., Heirman, W., & Stroobandt, D. (2011). Dynamic data folding with parameterizable FPGA configurations. ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS, 16(4).
Chicago author-date
Bruneel, Karel, Wim Heirman, and Dirk Stroobandt. 2011. “Dynamic Data Folding with Parameterizable FPGA Configurations.” Acm Transactions on Design Automation of Electronic Systems 16 (4).
Chicago author-date (all authors)
Bruneel, Karel, Wim Heirman, and Dirk Stroobandt. 2011. “Dynamic Data Folding with Parameterizable FPGA Configurations.” Acm Transactions on Design Automation of Electronic Systems 16 (4).
Vancouver
1.
Bruneel K, Heirman W, Stroobandt D. Dynamic data folding with parameterizable FPGA configurations. ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS. 2011;16(4).
IEEE
[1]
K. Bruneel, W. Heirman, and D. Stroobandt, “Dynamic data folding with parameterizable FPGA configurations,” ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS, vol. 16, no. 4, 2011.
@article{2019456,
  abstract     = {In many applications, subsequent data manipulations differ only in a small set of parameter values. Because of their reconfigurability, FPGAs (field programmable gate arrays) can be configured with a specialized circuit each time the parameter values change. This technique is called dynamic data folding. The specialized circuits are smaller and faster than their generic counterparts. However; the overhead involved in generating the configurations for the specialized circuits at runtime is very large when conventional tools are used, and this overhead will in many cases negate the benefit of using optimized configurations. This article introduces an automatic method for generating runtime parameterizable configurations from arbitrary Boolean circuits. These configurations, in which some of the configuration bits are expressed as a closed-form Boolean expression of a set of parameters, enable very fast run-time specialization, since specialization only involves evaluating these expressions. Our approach is validated on a ternary content-addressable memory (TCAM). We show that the specialized configurations, produced by our method use 2.82 times fewer LUTs than the generic configuration, and even 1.41 times fewer LUTs than the implementation generated by Xilinx Coregen. Moreover, while Coregen needs hand-crafted generators for each type of circuit, our toolflow can be applied to any VHDL design. Using our automatic and generally applicable method, run-time hardware optimization suddenly becomes feasible for a large class of applications.},
  articleno    = {43},
  author       = {Bruneel, Karel and Heirman, Wim and Stroobandt, Dirk},
  issn         = {1084-4309},
  journal      = {ACM TRANSACTIONS ON DESIGN AUTOMATION OF ELECTRONIC SYSTEMS},
  keywords     = {dynamic data folding,Automatic hardware synthesis,FPGA,run-time reconfiguration},
  language     = {eng},
  number       = {4},
  pages        = {29},
  title        = {Dynamic data folding with parameterizable FPGA configurations},
  url          = {http://dx.doi.org/10.1145/2003695.2003703},
  volume       = {16},
  year         = {2011},
}

Altmetric
View in Altmetric
Web of Science
Times cited: