Cost-aware prediction of uncorrected DRAM errors in the field
Document typeConference report
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Rights accessOpen Access
All rights reserved. This work is protected by the corresponding intellectual and industrial property rights. Without prejudice to any existing legal exemptions, reproduction, distribution, public communication or transformation of this work are prohibited without permission of the copyright holder
ProjectEuroEXA - Co-designed Innovation and System for Resilient Exascale Computing in Europe: From Applications to Silicon (EC-H2020-754337)
This paper presents and evaluates a method to predict DRAM uncorrected errors, a leading cause of hardware failures in large-scale HPC clusters. The method uses a random forest classifier, which was trained and evaluated using error logs from two years of production of the MareNostrum 3 supercomputer. By enabling the system to take measures to mitigate node failures, our method reduces lost compute time by up to 57%, a net saving of 21,000 node–hours per year. We release all source code as open source. We also discuss and clarify aspects of methodology that are essential for a DRAM prediction method to be useful in practice. We explain why standard evaluation metrics, such as precision and recall, are insufficient, and base the evaluation on a cost–benefit analysis. This methodology can help ensure that any DRAM error predictor is clear from training bias and has a clear cost–benefit calculation.
CitationBoixaderas, I. [et al.]. Cost-aware prediction of uncorrected DRAM errors in the field. A: International Conference for High Performance Computing, Networking, Storage and Analysis. "Proceedings of SC20: The International Conference for High Performance Computing, Networking, Storage and Analysis: Virtual Event, November 9-19, 2020". Institute of Electrical and Electronics Engineers (IEEE), 2020, p. 1-15. ISBN 978-1-7281-9998-6. DOI 10.1109/SC41405.2020.00065.