Evaluating the Impact of SDC on the GMRES Iterative Solver

Abstract

Increasing parallelism and transistor density, along with increasingly tighter energy and peak power constraints, may force exposure of occasionally incorrect computation or storage to application codes. Silent data corruption (SDC) will likely be infrequent, yet one SDC suffices to make numerical algorithms like iterative linear solvers cease progress towards the correct answer. Thus, we focus on resilience of the iterative linear solver GMRES to a single transient SDC. We derive inexpensive checks to detect the effects of an SDC in GMRES that work for a more general SDC model than presuming a bit flip. Our experiments show that when GMRES is used as the inner solver of an inner-outer iteration, it can "run through" SDC of almost any magnitude in the computationally intensive orthogonalization phase. That is, it gets the right answer using faulty data without any required roll back. Those SDCs which it cannot run through, get caught by our detection scheme.

References in corpus (1)

• Massively parallel solvers for elliptic PDEs in Numerical Weather- and Climate Prediction

Cited by in corpus (6)

• Improving Performance of Iterative Methods by Lossy Checkponting
• Exploiting Data Representation for Fault Tolerance
• Memory Vulnerability: A Case for Delaying Error Reporting
• Resilience in Numerical Methods: A Position on Fault Models and Methodologies
• Tolerating Silent Data Corruption in Opaque Preconditioners
• Soft Errors Detection and Automatic Recovery based on Replication combined with different Levels of Checkpointing