Soft Error Resilient Deep Learning Systems Using Neuron Gradient Statistics

Amarnath, Chandramouli; Mejri, Mohamed; Ma, Kwondo; Chatterjee, Abhijit

doi:10.1109/iolts56730.2022.9897815

Cited by 6 publications

(1 citation statement)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The second analysis we perform is on the domains the erroneous values belong to, to understand how faults affect the corrupted values in the output tensor and the error magnitude. This analysis can help investigate hardening solutions based on the expected values distribution, as in [23], [24]. In this perspective, it is essential to differentiate whether or not the erroneous values in the corrupted tensor fall within the nominal range.…”

Section: B Erroneous Values Domainsmentioning

confidence: 99%

Analyzing the Reliability of Alternative Convolution Implementations for Deep Learning Applications

Bolchini,

Cassano,

Miele

et al. 2023

2023 IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (DFT)

View full text Add to dashboard Cite

Convolution represents the core of Deep Learning (DL) applications, enabling the automatic extraction of features from raw input data. Several implementations of the convolution have been proposed. The impact of these different implementations on the performance of DL applications has been studied. However, no specific reliability-related analysis has been carried out. In this paper, we apply the CLASSES cross-layer reliability analysis methodology for an in-depth study aimed at: i) analyzing and characterizing the effects of Single Event Upsets occurring in Graphics Processing Units while executing the convolution operators; and ii) identifying whether a convolution implementation is more robust than others. The outcomes can then be exploited to tailor better hardening schemes for DL applications to improve reliability and reduce overhead.

show abstract

Section: B Erroneous Values Domainsmentioning

confidence: 99%