The adoption of Deep Neural Networks (DNNs) in several domains allows for increased effectiveness in applications that deal with massive data-intensive and complex data inputs. When employed in safety-critical scenarios, such as automotive, aerospace, healthcare, and autonomous robotics, assessing the DNNs' reliability and functional safety is crucial to ensure their correct in-field operation, even in the presence of hardware faults. However, the system complexity and the massive amounts of data to be processed by DNNs prevent the effective adoption of traditional strategies for reliability characterization and for identifying the most fault-sensitive structures. Accurate fault assessment strategies usually require unacceptable computational power and large evaluation times. On the other hand, faster strategies commonly lack accuracy in correctly representing system faults. Consequently, it is necessary to develop effective strategies that trade-off between performance and accuracy. This work analyses three reliability assessment strategies for deep neural networks and their underlying hardware, highlighting the main solutions and challenges in terms of evaluation performance and fault characterization accuracy. We overview different solutions to evaluate the hardware accelerators implementing DNNs at three abstraction levels: i) by physically injecting faults on a GPU running DNNs, ii) by performing microarchitectural characterization of GPUs to develop application-accurate error models, and iii) by using structure-aware cross-layer error modeling on DNN hardware accelerators.
Reliability Assessment of Large DNN Models:Trading Off Performance and Accuracy / Chen, Junchao; Esposito, Giuseppe; Fernandes dos Santos, Fernando; Guerrero-Balaguera, Juan-David; Kritikakou, Angeliki; Krstic, Milos; Limas-Sierra, Robert-Alexander; Rodriguez-Condia, Josie-Esteban; Sonza-Reorda, Matteo; Traiola, Marcello; Veronesi, Alessandro. - ELETTRONICO. - (2024), pp. 1-10. (Intervento presentato al convegno 2024 IFIP/IEEE 32nd International Conference on Very Large Scale Integration (VLSI-SoC) tenutosi a Tanger (MAR) nel 06-09 October 2024) [10.1109/VLSI-SoC62099.2024.10767814].
Reliability Assessment of Large DNN Models:Trading Off Performance and Accuracy
Esposito,Giuseppe;Guerrero-Balaguera,Juan-David;Limas-Sierra,Robert-Alexander;Rodriguez-Condia,Josie-Esteban;Sonza-Reorda,Matteo;
2024
Abstract
The adoption of Deep Neural Networks (DNNs) in several domains allows for increased effectiveness in applications that deal with massive data-intensive and complex data inputs. When employed in safety-critical scenarios, such as automotive, aerospace, healthcare, and autonomous robotics, assessing the DNNs' reliability and functional safety is crucial to ensure their correct in-field operation, even in the presence of hardware faults. However, the system complexity and the massive amounts of data to be processed by DNNs prevent the effective adoption of traditional strategies for reliability characterization and for identifying the most fault-sensitive structures. Accurate fault assessment strategies usually require unacceptable computational power and large evaluation times. On the other hand, faster strategies commonly lack accuracy in correctly representing system faults. Consequently, it is necessary to develop effective strategies that trade-off between performance and accuracy. This work analyses three reliability assessment strategies for deep neural networks and their underlying hardware, highlighting the main solutions and challenges in terms of evaluation performance and fault characterization accuracy. We overview different solutions to evaluate the hardware accelerators implementing DNNs at three abstraction levels: i) by physically injecting faults on a GPU running DNNs, ii) by performing microarchitectural characterization of GPUs to develop application-accurate error models, and iii) by using structure-aware cross-layer error modeling on DNN hardware accelerators.File | Dimensione | Formato | |
---|---|---|---|
Reliability_Assessment_of_Large_DNN_Models_Trading_Off_Performance_and_Accuracy.pdf
accesso riservato
Tipologia:
2a Post-print versione editoriale / Version of Record
Licenza:
Non Pubblico - Accesso privato/ristretto
Dimensione
1.55 MB
Formato
Adobe PDF
|
1.55 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/11583/2996512