Investigating the limits of AVF analysis in the presence of multiple bit errors

Mihalis Maniatakos, Maria K. Michael, Yiorgos Makris

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We investigate the complexity and utility of performing Multiple Bit Upset (MBU) vulnerability analysis in modern microprocessors. While the Single Bit Flip (SBF) model constitutes the prevailing mechanism for capturing the effect of Single Event Upsets (SEUs) due to alpha particle or neutron strikes in semiconductors, recent radiation studies in 90nm and 65nm technology nodes demonstrate that up to 55% of such strikes result in Multiple Bit Upsets (MBUs). Consequently, the accuracy of popular vulnerability analysis methods, such as the Architecural Vulnerability Factor (AVF) and Failures In Time (FIT) rate estimates based on the SBF assumption comes into question, especially in modern microprocessors which contain a significant amount of memory elements. Towards alleviating this concern, we present an extensive infrastructure which enables MBU vulnerability analysis in modern microprocessors. Using this infrastructure and a modern microprocessor model, we perform a large scale MBU vulnerability analysis study and we report two key findings: (i) the SBF fault model overestimates vulnerability by up to 71%, as compared to a more realistic modeling and distribution of faults in the 90nm and 65nm processes, and (ii) the rank-ordered lists of critical bits, as computed through the SBF and MBU models, respectively, are very similar, as indicated by the average rank difference of a bit which is less than 1.45%.

Original languageEnglish (US)
Title of host publicationProceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013
Pages49-54
Number of pages6
DOIs
StatePublished - Oct 15 2013
Event2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013 - Chania, Crete, Greece
Duration: Jul 8 2013Jul 10 2013

Other

Other2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013
CountryGreece
CityChania, Crete
Period7/8/137/10/13

Fingerprint

Factor analysis
Microprocessor chips
Alpha particles
Neutrons
Semiconductor materials
Radiation
Data storage equipment

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality

Cite this

Maniatakos, M., Michael, M. K., & Makris, Y. (2013). Investigating the limits of AVF analysis in the presence of multiple bit errors. In Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013 (pp. 49-54). [6604050] https://doi.org/10.1109/IOLTS.2013.6604050

Investigating the limits of AVF analysis in the presence of multiple bit errors. / Maniatakos, Mihalis; Michael, Maria K.; Makris, Yiorgos.

Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013. 2013. p. 49-54 6604050.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Maniatakos, M, Michael, MK & Makris, Y 2013, Investigating the limits of AVF analysis in the presence of multiple bit errors. in Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013., 6604050, pp. 49-54, 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013, Chania, Crete, Greece, 7/8/13. https://doi.org/10.1109/IOLTS.2013.6604050
Maniatakos M, Michael MK, Makris Y. Investigating the limits of AVF analysis in the presence of multiple bit errors. In Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013. 2013. p. 49-54. 6604050 https://doi.org/10.1109/IOLTS.2013.6604050
Maniatakos, Mihalis ; Michael, Maria K. ; Makris, Yiorgos. / Investigating the limits of AVF analysis in the presence of multiple bit errors. Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013. 2013. pp. 49-54
@inproceedings{61fe65ee5fe84dc88e40aeb44424494b,
title = "Investigating the limits of AVF analysis in the presence of multiple bit errors",
abstract = "We investigate the complexity and utility of performing Multiple Bit Upset (MBU) vulnerability analysis in modern microprocessors. While the Single Bit Flip (SBF) model constitutes the prevailing mechanism for capturing the effect of Single Event Upsets (SEUs) due to alpha particle or neutron strikes in semiconductors, recent radiation studies in 90nm and 65nm technology nodes demonstrate that up to 55{\%} of such strikes result in Multiple Bit Upsets (MBUs). Consequently, the accuracy of popular vulnerability analysis methods, such as the Architecural Vulnerability Factor (AVF) and Failures In Time (FIT) rate estimates based on the SBF assumption comes into question, especially in modern microprocessors which contain a significant amount of memory elements. Towards alleviating this concern, we present an extensive infrastructure which enables MBU vulnerability analysis in modern microprocessors. Using this infrastructure and a modern microprocessor model, we perform a large scale MBU vulnerability analysis study and we report two key findings: (i) the SBF fault model overestimates vulnerability by up to 71{\%}, as compared to a more realistic modeling and distribution of faults in the 90nm and 65nm processes, and (ii) the rank-ordered lists of critical bits, as computed through the SBF and MBU models, respectively, are very similar, as indicated by the average rank difference of a bit which is less than 1.45{\%}.",
author = "Mihalis Maniatakos and Michael, {Maria K.} and Yiorgos Makris",
year = "2013",
month = "10",
day = "15",
doi = "10.1109/IOLTS.2013.6604050",
language = "English (US)",
isbn = "9781479906628",
pages = "49--54",
booktitle = "Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013",

}

TY - GEN

T1 - Investigating the limits of AVF analysis in the presence of multiple bit errors

AU - Maniatakos, Mihalis

AU - Michael, Maria K.

AU - Makris, Yiorgos

PY - 2013/10/15

Y1 - 2013/10/15

N2 - We investigate the complexity and utility of performing Multiple Bit Upset (MBU) vulnerability analysis in modern microprocessors. While the Single Bit Flip (SBF) model constitutes the prevailing mechanism for capturing the effect of Single Event Upsets (SEUs) due to alpha particle or neutron strikes in semiconductors, recent radiation studies in 90nm and 65nm technology nodes demonstrate that up to 55% of such strikes result in Multiple Bit Upsets (MBUs). Consequently, the accuracy of popular vulnerability analysis methods, such as the Architecural Vulnerability Factor (AVF) and Failures In Time (FIT) rate estimates based on the SBF assumption comes into question, especially in modern microprocessors which contain a significant amount of memory elements. Towards alleviating this concern, we present an extensive infrastructure which enables MBU vulnerability analysis in modern microprocessors. Using this infrastructure and a modern microprocessor model, we perform a large scale MBU vulnerability analysis study and we report two key findings: (i) the SBF fault model overestimates vulnerability by up to 71%, as compared to a more realistic modeling and distribution of faults in the 90nm and 65nm processes, and (ii) the rank-ordered lists of critical bits, as computed through the SBF and MBU models, respectively, are very similar, as indicated by the average rank difference of a bit which is less than 1.45%.

AB - We investigate the complexity and utility of performing Multiple Bit Upset (MBU) vulnerability analysis in modern microprocessors. While the Single Bit Flip (SBF) model constitutes the prevailing mechanism for capturing the effect of Single Event Upsets (SEUs) due to alpha particle or neutron strikes in semiconductors, recent radiation studies in 90nm and 65nm technology nodes demonstrate that up to 55% of such strikes result in Multiple Bit Upsets (MBUs). Consequently, the accuracy of popular vulnerability analysis methods, such as the Architecural Vulnerability Factor (AVF) and Failures In Time (FIT) rate estimates based on the SBF assumption comes into question, especially in modern microprocessors which contain a significant amount of memory elements. Towards alleviating this concern, we present an extensive infrastructure which enables MBU vulnerability analysis in modern microprocessors. Using this infrastructure and a modern microprocessor model, we perform a large scale MBU vulnerability analysis study and we report two key findings: (i) the SBF fault model overestimates vulnerability by up to 71%, as compared to a more realistic modeling and distribution of faults in the 90nm and 65nm processes, and (ii) the rank-ordered lists of critical bits, as computed through the SBF and MBU models, respectively, are very similar, as indicated by the average rank difference of a bit which is less than 1.45%.

UR - http://www.scopus.com/inward/record.url?scp=84885207616&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84885207616&partnerID=8YFLogxK

U2 - 10.1109/IOLTS.2013.6604050

DO - 10.1109/IOLTS.2013.6604050

M3 - Conference contribution

AN - SCOPUS:84885207616

SN - 9781479906628

SP - 49

EP - 54

BT - Proceedings of the 2013 IEEE 19th International On-Line Testing Symposium, IOLTS 2013

ER -