Keywords (tags) and Publication List
Tao, Dingwen; Song, Shuaiwen Leon; Krishnamoorthy, Sriram; Wu, Panruo; Liang, Xin; Zhang, Eddy Z; Kerbyson, Darren; Chen, Zizhong New-Sum: A Novel Online ABFT Scheme For General Iterative Methods Conference Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC 2016), Association for Computing Machinery, Kyoto, Japan, 2016, ISBN: 9781450343145. Abstract | Links | BibTeX | Tags: Algorithm-based fault tolerance (abft), Checkpoint, Checksum, Iterative methods, Online error detection, Resilience, Rollback recovery, Silent data corruption (sdc)
2016
title = {New-Sum: A Novel Online ABFT Scheme For General Iterative Methods},
author = {Dingwen Tao and Shuaiwen Leon Song and Sriram Krishnamoorthy and Panruo Wu and Xin Liang and Eddy Z Zhang and Darren Kerbyson and Zizhong Chen},
url = {https://doi.org/10.1145/2907294.2907306},
doi = {10.1145/2907294.2907306},
isbn = {9781450343145},
year = {2016},
date = {2016-01-01},
booktitle = {Proceedings of the 25th ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC 2016)},
pages = {43–55},
publisher = {Association for Computing Machinery},
address = {Kyoto, Japan},
abstract = {Emerging high-performance computing platforms, with large component counts and lower power margins, are anticipated to be more susceptible to soft errors in both logic circuits and memory subsystems. We present an online algorithm-based fault tolerance (ABFT) approach to efficiently detect and recover soft errors for general iterative methods. We design a novel checksum-based encoding scheme for matrix-vector multiplication that is resilient to both arithmetic and memory errors. Our design decouples the checksum updating process from the actual computation, and allows adaptive checksum overhead control. Building on this new encoding mechanism, we propose two online ABFT designs that can effectively recover from errors when combined with a checkpoint/rollback scheme. These designs are capable of addressing scenarios under different error rates. Our ABFT approaches apply to a wide range of iterative solvers that primarily rely on matrix-vector multiplication and vector linear operations. We evaluate our designs through comprehensive analytical and empirical analysis. Experimental evaluation on the Stampede supercomputer demonstrates the low performance overheads incurred by our two ABFT schemes for preconditioned CG (0.4% and 2.2%) and preconditioned BiCGSTAB (1.0% and 4.0%) for the largest SPD matrix from UFL Sparse Matrix Collection. The evaluation also demonstrates the flexibility and effectiveness of our proposed designs for detecting and recovering various types of soft errors in general iterative methods.},
keywords = {Algorithm-based fault tolerance (abft), Checkpoint, Checksum, Iterative methods, Online error detection, Resilience, Rollback recovery, Silent data corruption (sdc)},
pubstate = {published},
tppubtype = {conference}
}