Arithmetic formats for implementing artificial neural networks on FPGAs

Abstract

This paper investigates the effect of arithmetic representation formats on the implementation of artificial neural networks (ANNs) on field-programmable gate arrays (FPGAs). The focus is on examining the tradeoffs between precision and range of various formats and the required FPGA resources. Basic ANN processing elements include multiplication and addition… (More)

Topics

22 Figures and Tables

Cite this paper

@article{Li2006ArithmeticFF, title={Arithmetic formats for implementing artificial neural networks on FPGAs}, author={Xiaoguang Li and Medhat A. Moussa and Shawki Areibi}, journal={Canadian Journal of Electrical and Computer Engineering}, year={2006}, volume={31}, pages={31-40} }