A Scalable Flexible SOM NoC-based Hardware Architecture

Abstract : In this paper, a parallel hardware implementation of a self-organizing map (SOM) is presented. Practical scalability and flexibility are the main architecture features which are obtained by using a Network-on-chip (NoC) approach for communication between neurons. The presented hardware architecture allows on-line learning and can be easily adapted for a large variety of applications without a considerable design effort. A hardware 5 × 5 SOM was validated through the FPGA implementation and its performances at a working frequency of 200MHz for a 32-element input vector reach 724 MCUPS in the learning and 1168 MCPS in the recall phase.
Complete list of metadatas

Cited literature [12 references]  Display  Hide  Download

https://hal.univ-lorraine.fr/hal-02065672
Contributor : Serge Weber <>
Submitted on : Tuesday, March 12, 2019 - 8:13:39 PM
Last modification on : Friday, March 15, 2019 - 1:28:40 AM
Long-term archiving on : Thursday, June 13, 2019 - 4:21:40 PM

File

wsom2016.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Mehdi Abadi, Slavisa Jovanovic, Khaled Ben Khalifa, Serge Weber, Mohamed Hédi Bedoui. A Scalable Flexible SOM NoC-based Hardware Architecture. Advances in Self-Organizing Maps and Learning Vector Quantization, pp.165-175, 2016, ⟨10.1007/978-3-319-28518-4_14⟩. ⟨hal-02065672⟩

Share

Metrics

Record views

14

Files downloads

34