In this paper we explore multiple and complementary approaches to analyze the single event upset susceptibility of a design implemented in Xilinx 28 nm SRAM-based FPGA. We choose as case study a neural network trained for a classification task. The techniques adopted are neutron irradiation, laser and emulated fault injection. These different techniques are complementary in the sense that they can be applied in different levels of the design integration, from the comprehensive but coarse reliability evaluation of the whole design, provided by the irradiation, to the fine grained and focused reliability investigation of individual modules or devices, provided by laser fault injection. Also, results from these different techniques can be compared allowing their use as crosschecking mechanisms.

Comparative Analysis of Inference Errors in a Neural Network Implemented in SRAM-Based FPGA Induced by Neutron Irradiation and Fault Injection Methods / Benevenuti, F.; Libano, F.; Pouget, V.; Kastensmidt, F. L.; Rech, P.. - (2018), pp. 1-6. (Intervento presentato al convegno 31st Symposium on Integrated Circuits and Systems Design, SBCCI 2018 tenutosi a bra nel 2018) [10.1109/SBCCI.2018.8533235].

Comparative Analysis of Inference Errors in a Neural Network Implemented in SRAM-Based FPGA Induced by Neutron Irradiation and Fault Injection Methods

Rech P.
2018-01-01

Abstract

In this paper we explore multiple and complementary approaches to analyze the single event upset susceptibility of a design implemented in Xilinx 28 nm SRAM-based FPGA. We choose as case study a neural network trained for a classification task. The techniques adopted are neutron irradiation, laser and emulated fault injection. These different techniques are complementary in the sense that they can be applied in different levels of the design integration, from the comprehensive but coarse reliability evaluation of the whole design, provided by the irradiation, to the fine grained and focused reliability investigation of individual modules or devices, provided by laser fault injection. Also, results from these different techniques can be compared allowing their use as crosschecking mechanisms.
2018
31st Symposium on Integrated Circuits and Systems Design, SBCCI 2018
USA
Institute of Electrical and Electronics Engineers Inc.
978-1-5386-7431-4
Benevenuti, F.; Libano, F.; Pouget, V.; Kastensmidt, F. L.; Rech, P.
Comparative Analysis of Inference Errors in a Neural Network Implemented in SRAM-Based FPGA Induced by Neutron Irradiation and Fault Injection Methods / Benevenuti, F.; Libano, F.; Pouget, V.; Kastensmidt, F. L.; Rech, P.. - (2018), pp. 1-6. (Intervento presentato al convegno 31st Symposium on Integrated Circuits and Systems Design, SBCCI 2018 tenutosi a bra nel 2018) [10.1109/SBCCI.2018.8533235].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11572/346623
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 17
  • ???jsp.display-item.citation.isi??? 4
social impact