FPDetect: Efficient Reasoning About Stencil Programs Using Selective Direct Evaluation

Abstract

We present FPDetect, a low overhead approach for detecting logical errors and soft errors affecting stencil computations without generating false positives. We develop an offline analysis that tightly estimates the number of floating-point bits preserved across stencil applications. This estimate rigorously bounds the values expected in the data space of the computation. Violations of this bound can be attributed with certainty to errors. FPDetect helps synthesize error detectors customized for user-specified levels of accuracy and coverage. FPDetect also enables overhead reduction techniques based on deploying these detectors coarsely in space and time. Experimental evaluations demonstrate the practicality of our approach.

arXiv commentAccepted in Journal of ACM Transactions on Architecture and Code Optimization

References in corpus (4)

• Tolerating Soft Errors in Processor Cores Using CLEAR (Cross-Layer Exploration for Architecting Resilience)
• Exploiting Data Representation for Fault Tolerance
• Resilience in Numerical Methods: A Position on Fault Models and Methodologies
• Reduced Precision Checking to Detect Errors in Floating Point Arithmetic