Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer

Sarah E. Michalak; Andrew J. DuBois; Curtis B. Storlie; Heather M. Quinn; William N. Rust; David H. DuBois; David G. Modl; Andrea Manuzzato; Sean P. Blanchard

doi:10.1109/TDMR.2012.2192736

Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer

Sarah E. Michalak, Andrew J. DuBois, Curtis B. Storlie, Heather M. Quinn, William N. Rust, David H. DuBois, David G. Modl, Andrea Manuzzato, Sean P. Blanchard

Quantitative Health Sciences

Research output: Contribution to journal › Article › peer-review

28 Scopus citations

Abstract

Microprocessor-based systems are a common design for high-performance computing (HPC) platforms. In these systems, several thousands of microprocessors can participate in a single calculation that may take weeks or months to complete. When used in this manner, a fault in any of the microprocessors could cause the computation to crash or cause silent data corruption (SDC), i.e., computationally incorrect results that originate from an undetected fault. In recent years, neutron-induced effects in HPC hardware have been observed, and researchers have started to study how neutrons impact microprocessor-based computations. This paper presents results from an accelerated neutron-beam test focusing on two microprocessors used in Roadrunner, which is the first petaflop supercomputer. Research questions of interest include whether the application running affects neutron susceptibility and whether different replicates of the hardware under test have different susceptibilities to neutrons. Estimated failures in time for crashes and for SDC are presented for the hardware under test, for the Triblade servers used for computation in Roadrunner, and for Roadrunner.

Original language	English (US)
Article number	6193419
Pages (from-to)	445-454
Number of pages	10
Journal	IEEE Transactions on Device and Materials Reliability
Volume	12
Issue number	2
DOIs	https://doi.org/10.1109/TDMR.2012.2192736
State	Published - 2012

Keywords

Failures in time (FIT)
neutron-beam testing
silent data corruption (SDC)
single-event effect
soft error

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Safety, Risk, Reliability and Quality
Electrical and Electronic Engineering

Access to Document

10.1109/TDMR.2012.2192736

Cite this

Michalak, S. E., DuBois, A. J., Storlie, C. B., Quinn, H. M., Rust, W. N., DuBois, D. H., Modl, D. G., Manuzzato, A., & Blanchard, S. P. (2012). Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer. IEEE Transactions on Device and Materials Reliability, 12(2), 445-454. Article 6193419. https://doi.org/10.1109/TDMR.2012.2192736

@article{67e826b8db7942e3a21f5033cfb81d20,

title = "Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer",

abstract = "Microprocessor-based systems are a common design for high-performance computing (HPC) platforms. In these systems, several thousands of microprocessors can participate in a single calculation that may take weeks or months to complete. When used in this manner, a fault in any of the microprocessors could cause the computation to crash or cause silent data corruption (SDC), i.e., computationally incorrect results that originate from an undetected fault. In recent years, neutron-induced effects in HPC hardware have been observed, and researchers have started to study how neutrons impact microprocessor-based computations. This paper presents results from an accelerated neutron-beam test focusing on two microprocessors used in Roadrunner, which is the first petaflop supercomputer. Research questions of interest include whether the application running affects neutron susceptibility and whether different replicates of the hardware under test have different susceptibilities to neutrons. Estimated failures in time for crashes and for SDC are presented for the hardware under test, for the Triblade servers used for computation in Roadrunner, and for Roadrunner.",

keywords = "Failures in time (FIT), neutron-beam testing, silent data corruption (SDC), single-event effect, soft error",

author = "Michalak, {Sarah E.} and DuBois, {Andrew J.} and Storlie, {Curtis B.} and Quinn, {Heather M.} and Rust, {William N.} and DuBois, {David H.} and Modl, {David G.} and Andrea Manuzzato and Blanchard, {Sean P.}",

note = "Funding Information: Manuscript received September 23, 2011; revised January 26, 2012; accepted February 27, 2012. Date of publication May 2, 2012; date of current version June 6, 2012. This work was supported by the U.S. Department of Energy. S. E. Michalak, C. B. Storlie, and W. N. Rust are with the Statistical Sciences Group, Los Alamos National Laboratory, Los Alamos, NM 87544 USA (e-mail: michalak@lanl.gov; storlie@lanl.gov; wnr@lanl.gov). A. J. DuBois, D. H. DuBois, D. G. Modl, and S. P. Blanchard are with the Systems Engineering and Integration Group, Los Alamos National Laboratory, Los Alamos, NM 87545 USA (e-mail: ajd@lanl.gov; dhd@lanl.gov; digem@lanl.gov; seanb@lanl.gov). H. M. Quinn is with the Space Data Systems Group, Los Alamos National Laboratory, Los Alamos, NM 87545 USA (e-mail: hquinn@lanl.gov). A. Manuzzato is with the University of Milan, Milan, Italy (e-mail: andrea.manuzzato@ieee.org). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TDMR.2012.2192736",

year = "2012",

doi = "10.1109/TDMR.2012.2192736",

language = "English (US)",

volume = "12",

pages = "445--454",

journal = "IEEE Transactions on Device and Materials Reliability",

issn = "1530-4388",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "2",

}

TY - JOUR

T1 - Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer

AU - Michalak, Sarah E.

AU - DuBois, Andrew J.

AU - Storlie, Curtis B.

AU - Quinn, Heather M.

AU - Rust, William N.

AU - DuBois, David H.

AU - Modl, David G.

AU - Manuzzato, Andrea

AU - Blanchard, Sean P.

N1 - Funding Information: Manuscript received September 23, 2011; revised January 26, 2012; accepted February 27, 2012. Date of publication May 2, 2012; date of current version June 6, 2012. This work was supported by the U.S. Department of Energy. S. E. Michalak, C. B. Storlie, and W. N. Rust are with the Statistical Sciences Group, Los Alamos National Laboratory, Los Alamos, NM 87544 USA (e-mail: michalak@lanl.gov; storlie@lanl.gov; wnr@lanl.gov). A. J. DuBois, D. H. DuBois, D. G. Modl, and S. P. Blanchard are with the Systems Engineering and Integration Group, Los Alamos National Laboratory, Los Alamos, NM 87545 USA (e-mail: ajd@lanl.gov; dhd@lanl.gov; digem@lanl.gov; seanb@lanl.gov). H. M. Quinn is with the Space Data Systems Group, Los Alamos National Laboratory, Los Alamos, NM 87545 USA (e-mail: hquinn@lanl.gov). A. Manuzzato is with the University of Milan, Milan, Italy (e-mail: andrea.manuzzato@ieee.org). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TDMR.2012.2192736

PY - 2012

Y1 - 2012

N2 - Microprocessor-based systems are a common design for high-performance computing (HPC) platforms. In these systems, several thousands of microprocessors can participate in a single calculation that may take weeks or months to complete. When used in this manner, a fault in any of the microprocessors could cause the computation to crash or cause silent data corruption (SDC), i.e., computationally incorrect results that originate from an undetected fault. In recent years, neutron-induced effects in HPC hardware have been observed, and researchers have started to study how neutrons impact microprocessor-based computations. This paper presents results from an accelerated neutron-beam test focusing on two microprocessors used in Roadrunner, which is the first petaflop supercomputer. Research questions of interest include whether the application running affects neutron susceptibility and whether different replicates of the hardware under test have different susceptibilities to neutrons. Estimated failures in time for crashes and for SDC are presented for the hardware under test, for the Triblade servers used for computation in Roadrunner, and for Roadrunner.

AB - Microprocessor-based systems are a common design for high-performance computing (HPC) platforms. In these systems, several thousands of microprocessors can participate in a single calculation that may take weeks or months to complete. When used in this manner, a fault in any of the microprocessors could cause the computation to crash or cause silent data corruption (SDC), i.e., computationally incorrect results that originate from an undetected fault. In recent years, neutron-induced effects in HPC hardware have been observed, and researchers have started to study how neutrons impact microprocessor-based computations. This paper presents results from an accelerated neutron-beam test focusing on two microprocessors used in Roadrunner, which is the first petaflop supercomputer. Research questions of interest include whether the application running affects neutron susceptibility and whether different replicates of the hardware under test have different susceptibilities to neutrons. Estimated failures in time for crashes and for SDC are presented for the hardware under test, for the Triblade servers used for computation in Roadrunner, and for Roadrunner.

KW - Failures in time (FIT)

KW - neutron-beam testing

KW - silent data corruption (SDC)

KW - single-event effect

KW - soft error

UR - http://www.scopus.com/inward/record.url?scp=84862004001&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84862004001&partnerID=8YFLogxK

U2 - 10.1109/TDMR.2012.2192736

DO - 10.1109/TDMR.2012.2192736

M3 - Article

AN - SCOPUS:84862004001

SN - 1530-4388

VL - 12

SP - 445

EP - 454

JO - IEEE Transactions on Device and Materials Reliability

JF - IEEE Transactions on Device and Materials Reliability

IS - 2

M1 - 6193419

ER -

Assessment of the impact of cosmic-ray-induced neutrons on hardware in the roadrunner supercomputer

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this