Understanding time-varying vulnerability accross GPU Program Lifetime

Hao Qiu; Semiu A. Olowogemo; Bor Tyng Lin; William H. Robinson; Daniel B. Limbrick

doi:10.1109/dft56152.2022.9962365

Understanding time-varying vulnerability accross GPU Program Lifetime

Hao Qiu
, Semiu A. Olowogemo
, Bor Tyng Lin
, William H. Robinson
, Daniel B. Limbrick

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Time-varying behaviors of GPU program vulnerability could be exploited to reduce overheads for fault-tolerant designs. However, the inherent parallelism and performance overheads for massive fault injection (FI) hindered such assessments using FI. NVBitFI, a GPU FI tool featuring high-performance and good compatibility, allows time-varying vulnerability evaluations using FI within a reasonable time. We extended NVBitFI to control FI tests on the temporal dimension. A scalable workflow characterizing the time-varying vulnerability of GPU programs at two granularities is presented. A convenient approach to profile vulnerability with actual GPU time is also proposed. Results obtained from 60K fault injections demonstrated the feasibility of the proposed methodologies. A case study exploring the improved instruction-level grouping is presented. More than 340K faults are injected into the vectorAdd kernel to show the possibility to generalize the time-varying behavior of smaller inputs to realistic workloads with large inputs.

Original language	English
Title of host publication	35th IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2022
Volume	2022-
DOIs	https://doi.org/10.1109/dft56152.2022.9962365
State	Published - 2022

Access to Document

10.1109/dft56152.2022.9962365

Cite this

@inproceedings{33260b1c982d4e8cad4576118b3af9a1,

title = "Understanding time-varying vulnerability accross GPU Program Lifetime",

abstract = "Time-varying behaviors of GPU program vulnerability could be exploited to reduce overheads for fault-tolerant designs. However, the inherent parallelism and performance overheads for massive fault injection (FI) hindered such assessments using FI. NVBitFI, a GPU FI tool featuring high-performance and good compatibility, allows time-varying vulnerability evaluations using FI within a reasonable time. We extended NVBitFI to control FI tests on the temporal dimension. A scalable workflow characterizing the time-varying vulnerability of GPU programs at two granularities is presented. A convenient approach to profile vulnerability with actual GPU time is also proposed. Results obtained from 60K fault injections demonstrated the feasibility of the proposed methodologies. A case study exploring the improved instruction-level grouping is presented. More than 340K faults are injected into the vectorAdd kernel to show the possibility to generalize the time-varying behavior of smaller inputs to realistic workloads with large inputs.",

author = "Hao Qiu and Olowogemo, \{Semiu A.\} and Lin, \{Bor Tyng\} and Robinson, \{William H.\} and Limbrick, \{Daniel B.\}",

year = "2022",

doi = "10.1109/dft56152.2022.9962365",

language = "English",

volume = "2022-",

booktitle = "35th IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2022",

}

TY - GEN

T1 - Understanding time-varying vulnerability accross GPU Program Lifetime

AU - Qiu, Hao

AU - Olowogemo, Semiu A.

AU - Lin, Bor Tyng

AU - Robinson, William H.

AU - Limbrick, Daniel B.

PY - 2022

Y1 - 2022

N2 - Time-varying behaviors of GPU program vulnerability could be exploited to reduce overheads for fault-tolerant designs. However, the inherent parallelism and performance overheads for massive fault injection (FI) hindered such assessments using FI. NVBitFI, a GPU FI tool featuring high-performance and good compatibility, allows time-varying vulnerability evaluations using FI within a reasonable time. We extended NVBitFI to control FI tests on the temporal dimension. A scalable workflow characterizing the time-varying vulnerability of GPU programs at two granularities is presented. A convenient approach to profile vulnerability with actual GPU time is also proposed. Results obtained from 60K fault injections demonstrated the feasibility of the proposed methodologies. A case study exploring the improved instruction-level grouping is presented. More than 340K faults are injected into the vectorAdd kernel to show the possibility to generalize the time-varying behavior of smaller inputs to realistic workloads with large inputs.

AB - Time-varying behaviors of GPU program vulnerability could be exploited to reduce overheads for fault-tolerant designs. However, the inherent parallelism and performance overheads for massive fault injection (FI) hindered such assessments using FI. NVBitFI, a GPU FI tool featuring high-performance and good compatibility, allows time-varying vulnerability evaluations using FI within a reasonable time. We extended NVBitFI to control FI tests on the temporal dimension. A scalable workflow characterizing the time-varying vulnerability of GPU programs at two granularities is presented. A convenient approach to profile vulnerability with actual GPU time is also proposed. Results obtained from 60K fault injections demonstrated the feasibility of the proposed methodologies. A case study exploring the improved instruction-level grouping is presented. More than 340K faults are injected into the vectorAdd kernel to show the possibility to generalize the time-varying behavior of smaller inputs to realistic workloads with large inputs.

UR - https://dx.doi.org/10.1109/DFT56152.2022.9962365

U2 - 10.1109/dft56152.2022.9962365

DO - 10.1109/dft56152.2022.9962365

M3 - Conference contribution

VL - 2022-

BT - 35th IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems, DFT 2022

ER -

Understanding time-varying vulnerability accross GPU Program Lifetime

Abstract

Access to Document

Other files and links

Fingerprint

Cite this