Exportar Publicação

John, L. K., França, F. M. G., Mitra, S., Susskind, Z., Lima, P. M. V., Miranda, I. D. S....Breternitz Jr., M. (2023). Dendrite-inspired computing to improve resilience of neural networks to faults in emerging memory technologies. In 2023 IEEE International Conference on Rebooting Computing (ICRC). San Diego, CA, USA : IEEE.

L. JOHN et al.,  "Dendrite-inspired computing to improve resilience of neural networks to faults in emerging memory technologies", in 2023 IEEE Int. Conf. on Rebooting Computing (ICRC), San Diego, CA, USA , IEEE, 2023

@inproceedings{john2023_1715473524050,
	author = "John, L. K. and França, F. M. G. and Mitra, S. and Susskind, Z. and Lima, P. M. V. and Miranda, I. D. S. and John, E. B. and Dutra, D. L. C. and Breternitz Jr., M.",
	title = "Dendrite-inspired computing to improve resilience of neural networks to faults in emerging memory technologies",
	booktitle = "2023 IEEE International Conference on Rebooting Computing (ICRC)",
	year = "2023",
	editor = "",
	volume = "",
	number = "",
	series = "",
	doi = "10.1109/ICRC60800.2023.10386729",
	publisher = "IEEE",
	address = "San Diego, CA, USA ",
	organization = "",
	url = "https://ieeexplore.ieee.org/xpl/conhome/10385233/proceeding"
}

TY  - CPAPER
TI  - Dendrite-inspired computing to improve resilience of neural networks to faults in emerging memory technologies
T2  - 2023 IEEE International Conference on Rebooting Computing (ICRC)
AU  - John, L. K.
AU  - França, F. M. G.
AU  - Mitra, S.
AU  - Susskind, Z.
AU  - Lima, P. M. V.
AU  - Miranda, I. D. S.
AU  - John, E. B.
AU  - Dutra, D. L. C.
AU  - Breternitz Jr., M.
PY  - 2023
DO  - 10.1109/ICRC60800.2023.10386729
CY  - San Diego, CA, USA 
UR  - https://ieeexplore.ieee.org/xpl/conhome/10385233/proceeding
AB  - Mimicking biological neurons by focusing on the excitatory/inhibitory decoding performed by dendritic trees offers an intriguing alternative to the traditional integrate-and-fire McCullogh-Pitts neuron stylization. Weightless Neural Networks (WNN), which rely on value lookups from tables, emulate the integration process in dendrites and have demonstrated notable advantages in terms of energy efficiency. In this paper, we delve into the WNN paradigm from the perspective of reliability and fault tolerance. Through a series of fault injection experiments, we illustrate that WNNs exhibit remarkable resilience to both transient (soft) errors and permanent faults. Notably, WNN models experience minimal deterioration in accuracy even when subjected to fault rates of up to 5%. This resilience makes them well-suited for implementation in emerging memory technologies for binary or multiple bits-per-cell storage with reduced reliance on memory block-level error resilience features. By offering a novel perspective on neural network modeling and highlighting the robustness of WNNs, this research contributes to the broader understanding of fault tolerance in neural networks, particularly in the context of emerging memory technologies.
ER  -