FPNA: applications and implementations

Bernard Girau 1
1 CORTEX - Neuromimetic intelligence
INRIA Lorraine, LORIA - Laboratoire Lorrain de Recherche en Informatique et ses Applications
Abstract : Neural networks are usually considered as naturally parallel computing models. But the number of operators and the complex connection graph of standard neural models can not be handled by digital hardware devices. The Field Programmable Neural Arrays (FPNA) framework introduced in chapter "FPNA: concepts and properties" reconciles simple hardware topologies with complex neural architectures, thanks to some configurable hardware principles applied to neural computation. This two-chapter study (1- concepts and properties, 2- applications and implementations) gathers the different results that have been published about the FPNA concept, as well as some unpublished ones. This second part shows how FPNAs lead to powerful neural architectures that are easy to map onto digital hardware: applications and implementations are described, focusing on a class of synchronous FPNA-derived neural networks, for which on-chip learning is also available.
Type de document :
Chapitre d'ouvrage
Amos Omondi, Jagath Rajapakse. FPGA Implementations of Neural Networks, Kluwer Academic Publishers, pp.43-79, 2004
Liste complète des métadonnées

https://hal.inria.fr/inria-00100073
Contributeur : Publications Loria <>
Soumis le : mardi 26 septembre 2006 - 10:13:52
Dernière modification le : jeudi 11 janvier 2018 - 06:19:48

Identifiants

  • HAL Id : inria-00100073, version 1

Collections

Citation

Bernard Girau. FPNA: applications and implementations. Amos Omondi, Jagath Rajapakse. FPGA Implementations of Neural Networks, Kluwer Academic Publishers, pp.43-79, 2004. 〈inria-00100073〉

Partager

Métriques

Consultations de la notice

184