Abstract
This paper is focused on hardware implementation of neural networks. We propose a reconfigurable, low-cost and readily available hardware architecture for an artificial neuron. For this purpose, we use field-programmable gate arrays i.e. FPGAs. As the state-of-the-art FPGAs still lack the gate density necessary to the implementation of large neural networks of thousands of neurons, we use a stochastic process to implement efficiently the computation performed by a neuron. This paper describes and compares the characteristics of two architectures designed to implement feed-forward fully connected artificial neural networks: the first FPGA prototype is based on traditional adders and multipliers of binary inputs while the second takes advantage of stochastic representation of the inputs. The paper compares both prototypes using the time × area classic factor.
Similar content being viewed by others
References
Bade SL, Hutchings BL (1994) FPGA-based stochastic neural networks—implementation, IEEE workshop on FPGAs for custom computing machines, Napa
Brown BD, Card HC (2001) Stochastic neural computation I: computational elements. IEEE Trans Comp 50(9):891–905
Brown BD, Card HC (2001), Stochastic neural computation II: soft competitive learning. IEEE Trans Comp 50(9):906–920
Daalen MV, Jeavons P, Shawe-Taylor J (1993) A stochastic neural architecture that exploits dynamically reconfigurable FPGAs. In: Proceedings of the IEEE workshop on FPGAs for custom computing machines
Daalen MV, Jeavons P, Shawe-Taylor J (1993) A device for generating binary sequence for stochastic computing. Electron Lett 29(1):80–81
Gaines BR (1969) Stochastic computing systems. Adv Inf Syst Sci 2:37–172
Hassoun MH (1995) Fundamentals of artificial neural networks. MIT Press, Cambridge
Moerland P, Fiesler E (1996) Neural network adaptation to hardware implementations. In: Fiesler E, Beale R (eds) Handbook of neural computation
Navabi Z (1998) VHDL-Analysis and modelling of digital systems, 2nd edn. McGraw Hill
Xilinx Inc., http://www.xilinx.com
Acknowledgment
We are grateful to the reviewers and the editor that contributed to the great improvement of the original version of this paper with their valuable comments and suggestions. We also are thankful to FAPERJ (Fundação de Amparo à Pesquisa do Estado do Rio de janeiro, http://www.faperj.br) and CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico, http://www.cnpq.br) for their continuous financial support.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Nedjah, N., de Macedo Mourelle, L. Reconfigurable hardware for neural networks: binary versus stochastic. Neural Comput & Applic 16, 249–255 (2007). https://doi.org/10.1007/s00521-007-0086-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-007-0086-x