SAFIIRA: A heavy-ion multi-purpose irradiation facility in Brazil

Aguiar, Vitor A. P.; Medina, N. H.; Added, N.; Macchione, Eduardo Luiz Augusto; Alberton, S. G.; Leite, Alisson Rodolfo; Aguirre, Fernando; Ribas, Roberto Vicençotto; Perego, Claudio; Fagundes, Ligia; Terassi, João Carlos; Brage, J.A.P.; Simões, R. F.; Morais, Otavio Benedito de; Almeida, Eduardo Alves de; Joaquim, Pedro Manoel; Souza, Matheus S.; Cecotte, Alex Fernando Monego; Martins, Renato; Duarte, J. L. M.; Scarduelli, V.; Allegro, P.R.P.; Escudeiro, R.; Leistenschneider, E.; Oliveira, R. A. N.; Servelo, W. A.; Silva, M. T.; Sarmento, V. E.; Carreira, C. A.; Abreu, Juliana; Silva, S. C.; Santos, Hadassa Campos; Rodrigues, C. L.; Assis, R.F.; Silva, Tatiane da Franca; Tabacniks, M.H.; Joaquim, A. S.; Minas, J. H. P.; Kashinsky, D.A.; Guazzelli, Marcilei A.; Seixas, L. E.; Finco, Saulo; Benevenutti, F.

doi:10.1063/1.5138644

Cited by 10 publications

(2 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The heavy ions accelerated irradiation experiments were performed using the SAFIIRA [20] test setup dedicated to electronic devices qualification at the tandem electrostatic 8 MeV Pelletron accelerator of the LAFN-USP facility, Brazil. In this setup, the beam is produced by multiple collimations, defocusing and scattering by thin gold foils, achieving high uniformity, large area and very low particle fluxes in the range from 10 2 to 10 5 particles/cm 2 /s.…”

Section: B Beam Testing Campaignmentioning

confidence: 99%

Robust Convolutional Neural Networks in SRAM-based FPGAs: a Case Study in Image Classification

Benevenuti

Kastensmidt

Oliveira

et al. 2021

JICS

Self Cite

View full text Add to dashboard Cite

This work discusses the main aspects of vulnerability and degradation of accuracy of an image classification engine implemented into SRAM-based FPGAs under faults. The image classification engine is an all-convolutional neural-network (CNN) trained with a dataset of traffic sign recognition benchmark. The Caffe and Ristretto frameworks were used for CNN training and fine-tuning while the ZynqNet inference engine was adopted as hardware implementation on a Xilinx 28 nm SRAM-based FPGA. The CNN under test was generated using an evolutive approach based on genetic algorithm. The methodologies for qualifying this CNN under faults is presented and both heavy-ions accelerated irradiation and emulated fault injection were performed. To cross validate results from radiation and fault injection, different implementations of the same CNN were tested using reduced arithmetic precision and protection of user data by Hamming codes, in combination with configuration memory healing by the scrubbing mechanism available in Xilinx FPGA. Some of these alternative implementations increased significantly the mission time of the CNN, when compared to the original ZynqNet operating on 32 bits floating point number, and the experiment suggests areas for further improvements on the fault injection methodology in use.

show abstract

Section: B Beam Testing Campaignmentioning

confidence: 99%

Robust Convolutional Neural Networks in SRAM-based FPGAs: a Case Study in Image Classification

Benevenuti

Kastensmidt

Oliveira

et al. 2021

JICS

Self Cite

View full text Add to dashboard Cite

show abstract

“…Two heavy ion irradition campains were performed on the fault tolerant DAS at the Laboratório Aberto de Física Nuclear at the Universidade de São Paulo (LAFN-USP), Brazil [31], with ion beams produced and accelerated by the São Paulo 8UD Pelletron Accelerator, generating a 16 O ion beam with 22 µm penetration into the silicon. The heavy-ion beam was accelerated up to 36 MeV and reduced its intensity to hundred particles/s/cm² (as recommended by the European Space Agency (ESA) for SEU tests [32]) using magnetic defocusing techniques and two thin gold foils (near 1 mg/cm²) in the SAFIIRA system [33]. This new system was built to study heavy-ion beam effects in electronic components.…”

Section: A Test Setupmentioning

confidence: 99%

Testing a Fault Tolerant Mixed-Signal Design Under TID and Heavy Ions

González

Machado

Vaz³

et al. 2021

JICS

View full text Add to dashboard Cite

This work presents results of three distinct radiation tests performed upon a fault-tolerant data acquisition system comprising a design diversity redundancy technique. The first and second experiments are Total Ionizing Dose (TID) essays, comprising gamma and X-ray irradiations. The last experiment considers single event effects, in which two heavy ion irradiation campaigns are carried out. The case study system comprises three analog-to-digital converters and two software-based voters, besides additional software and hardware resources used for controlling, monitoring and memory management. The applied Diversity Triple Modular Redundancy (DTMR) technique, comprises different levels of diversity (temporal and architectural). The circuit was designed in a programmable System-on-Chip (PSoC), fabricated in a 130nm CMOS technology process. Results show that the technique may increase the lifetime of the system under TID if comparing with a non-redundant implementation. Considering the heavy ions experiments the system was proved effective to tolerate 100% of the observed errors originated in the converters, while errors in the processing unit present a higher criticality. Critical errors occurring in one of the voters were also observed. A second heavy-ion campaign was then carried out to investigate the voters reliability, comparing the dynamic cross-section of three different software-based voter schemes implemented in the considered PSoC.

show abstract