On-Line Testing of Neuromorphic Hardware - Archive ouverte HAL
Conference Papers Year : 2023

On-Line Testing of Neuromorphic Hardware

Abstract

We propose an on-line testing methodology for neuromorphic hardware supporting spiking neural networks. Testing aims at detecting in real-time abnormal operation due to hardware-level faults, as well as screening of outlier or corner inputs that are prone to misprediction. Testing is enabled by two on-chip classifiers that prognosticate, based on a low-dimensional set of features extracted with spike counting, whether the network will make a correct prediction. The system of classifiers is capable of evaluating the confidence of the decision, and when the confidence is judged low a replay operation helps to resolve the ambiguity. The testing methodology is demonstrated by fully embedding it in a custom FPGA-based neuromorphic hardware platform. It operates in the background being totally nonintrusive to the network operation, while offering a zero-latency test decision for the vast majority of inferences.
Fichier principal
Vignette du fichier
On_Line_Testing_of_Neuromorphic_Hardware.pdf (1 Mo) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-04060634 , version 1 (06-04-2023)

Identifiers

Cite

Theofilos Spyrou, Haralampos-G. Stratigopoulos. On-Line Testing of Neuromorphic Hardware. 2023 IEEE European Test Symposium (ETS), May 2023, Venise, Italy. pp.1-6, ⟨10.1109/ETS56758.2023.10174077⟩. ⟨hal-04060634⟩
123 View
169 Download

Altmetric

Share

More