Testing and Reliability of Spiking Neural Networks: A Review of the State-of-the-Art

Haralampos-G. Stratigopoulos (1) , Theofilos Spyrou (1) , Spyridon Raptis (1)

1 CIAN - Circuits Intégrés Numériques et Analogiques

Neuromorphic computing based on Spiking Neural Networks (SNNs) is an emerging computing paradigm inspired by the functionality of the biological brain. Given its potential to revolutionize the power efficiency of many Artificial Intelligence (AI) applications, heavy research is underway on algorithms, hardware implementations, and applications. This article focuses on testing and reliability of hardware implementations providing a review of the state-of-the-art.

Mots clés: Neuromorphic computing; Spiking neural networks; Testing; Reliabilility; Dependabilility

Domaines: Intelligence artificielle [cs.AI]; Micro et nanotechnologies/Microélectronique

Haralampos-G. Stratigopoulos, Theofilos Spyrou, Spyridon Raptis. Testing and Reliability of Spiking Neural Networks: A Review of the State-of-the-Art. 36th IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT 2023), Oct 2023, Juan-Les-Pins, France. ⟨hal-04176109⟩