WORTEX: Worst-Case Execution Time and Energy Estimation in Low-Power Microprocessors Using Explainable ML

Authors Hugo Reymond , Abderaouf Nassim Amalou , Isabelle Puaut



PDF
Thumbnail PDF

File

OASIcs.WCET.2024.1.pdf
  • Filesize: 1.18 MB
  • 14 pages

Document Identifiers

Author Details

Hugo Reymond
  • Univ. Rennes, INRIA, CNRS, IRISA, Rennes, France
Abderaouf Nassim Amalou
  • Université Paris-Saclay, CEA, List, F-91120, Palaiseau, France
Isabelle Puaut
  • Univ. Rennes, INRIA, CNRS, IRISA, Rennes, France

Acknowledgements

The authors would like to thank Hector Chabot for his essential work on the creation of this paper dataset.

Cite As Get BibTex

Hugo Reymond, Abderaouf Nassim Amalou, and Isabelle Puaut. WORTEX: Worst-Case Execution Time and Energy Estimation in Low-Power Microprocessors Using Explainable ML. In 22nd International Workshop on Worst-Case Execution Time Analysis (WCET 2024). Open Access Series in Informatics (OASIcs), Volume 121, pp. 1:1-1:14, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2024) https://doi.org/10.4230/OASIcs.WCET.2024.1

Abstract

Real-time and energy-constrained systems heavily rely on estimates of the worst-case execution time (WCET) and worst-case energy consumption (WCEC) of code snippets to ensure trustworthy operation. Designing architecture-specific analytical models for time and energy is often challenging and time-consuming. In situations where analytical models are unavailable or incomplete, machine learning (ML) techniques emerge as a promising solution to build WCEC/WCET models. This paper introduces WORTEX, a toolkit for WCEC/WCET estimation of basic blocks based on ML techniques. To ensure the real-world applicability of its models, WORTEX extracts large datasets of basic blocks from real programs and precisely measures their energy consumption/execution time on the physical target platform. The dataset is used to train various WCEC/WCET models using different ML techniques. Experimental results on simple and time-predictable hardware show that even the most basic ML techniques provide accurate results, that never underestimate actual values. We also discuss the use of explainability techniques to gain trustworthiness for the models.

Subject Classification

ACM Subject Classification
  • Computer systems organization → Real-time systems
Keywords
  • Worst-Case Execution Time (WCET)
  • Worst-Case Energy Consumption (WCEC)
  • Machine Learning
  • Explainable ML models

Metrics

  • Access Statistics
  • Total Accesses (updated on a weekly basis)
    0
    PDF Downloads

References

  1. Peter Altenbernd, Jan Gustafsson, Björn Lisper, and Friedhelm Stappert. Early execution time-estimation through automatically generated timing models. Real-Time Systems, 2016. URL: https://doi.org/10.1007/s11241-016-9250-7.
  2. Abderaouf Nassim Amalou, Elisa Fromont, and Isabelle Puaut. CAWET: Context-Aware Worst-Case Execution Time Estimation Using Transformers. In 35th Euromicro Conference on Real-Time Systems (ECRTS 2023), 2023. URL: https://doi.org/10.4230/LIPIcs.ECRTS.2023.7.
  3. Abderaouf Nassim Amalou, Isabelle Puaut, and Gilles Muller. WE-HML: hybrid WCET estimation using machine learning for architectures with caches. In IEEE 27th International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), 2021. URL: https://doi.org/10.1109/RTCSA52859.2021.00011.
  4. Gautier Berthou, Kevin Marquet, Tanguy Risset, and Guillaume Salagnac. Accurate power consumption evaluation for peripherals in ultra low-power embedded systems. In Global Internet of Things Summit (GIoTS), 2020. URL: https://doi.org/10.1109/GIOTS49054.2020.9119593.
  5. Armelle Bonenfant, Denis Claraz, Marianne De Michiel, and Pascal Sotin. Early WCET prediction using machine learning. In 17th International Workshop on Worst-Case Execution Time Analysis (WCET), 2017. URL: https://doi.org/10.4230/OASIcs.WCET.2017.5.
  6. Jongouk Choi, Larry Kittinger, Qingrui Liu, and Changhee Jung. Compiler-directed high-performance intermittent computation with power failure immunity. In 28th IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS'22), 2022. URL: https://doi.org/10.1109/RTAS54340.2022.00012.
  7. Anderson Faustino Da Silva, Bruno Conde Kind, José Wesley de Souza Magalhães, Jerônimo Nunes Rocha, Breno Campos Ferreira Guimaraes, and Fernando Magno Quinão Pereira. Anghabench: A suite with one million compilable c benchmarks for code-size reduction. In IEEE/ACM International Symposium on Code Generation and Optimization (CGO), 2021. URL: https://doi.org/10.1109/CGO51591.2021.9370322.
  8. Eva García-Martín, Crefeda Faviola Rodrigues, Graham Riley, and Håkan Grahn. Estimation of energy consumption in machine learning. Journal of Parallel and Distributed Computing, 2019. URL: https://doi.org/10.1016/j.jpdc.2019.07.007.
  9. Jan Gustafsson, Adam Betts, Andreas Ermedahl, and Björn Lisper. The Mälardalen WCET benchmarks: Past, present and future. In 10th International Workshop on Worst-Case Execution Time Analysis (WCET), 2010. URL: https://doi.org/10.4230/OASIcs.WCET.2010.136.
  10. Damien Hardy, Benjamin Rouxel, and Isabelle Puaut. The heptane static worst-case execution time estimation tool. In 17th International Workshop on Worst-Case Execution Time Analysis (WCET), 2017. URL: https://doi.org/10.4230/OASIcs.WCET.2017.8.
  11. Thomas Huybrechts, Siegfried Mercelis, and Peter Hellinckx. A new hybrid approach on WCET analysis for real-time systems using machine learning. In 18th International Workshop on Worst-Case Execution Time Analysis (WCET), 2018. URL: https://doi.org/10.4230/OASIcs.WCET.2018.5.
  12. Thomas Huybrechts, Philippe Reiter, Siegfried Mercelis, Jeroen Famaey, Steven Latré, and Peter Hellinckx. Automated testbench for hybrid machine learning-based worst-case energy consumption analysis on batteryless iot devices. Energies, 2021. URL: https://doi.org/10.3390/en14133914.
  13. R. Jayaseelan, T. Mitra, and X. Li. Estimating the worst-case energy consumption of embedded software. In 2013 IEEE 19th Real-Time and Embedded Technology and Applications Symposium (RTAS), April 2006. URL: https://doi.org/10.1109/RTAS.2006.17.
  14. KEYSIGHT. N6700 Modular Power System Family. URL: https://www.keysight.com/fr/en/assets/7018-05443/data-sheets/5992-1857.pdf.
  15. Roger Koenker. Quantile regression, volume 38. Cambridge university press, 2005. Google Scholar
  16. Vikash Kumar. Deep neural network approach to estimate early worst-case execution time. In IEEE/AIAA 40th Digital Avionics Systems Conference (DASC), 2021. URL: https://doi.org/10.1109/DASC52595.2021.9594326.
  17. Vikash Kumar. Estimation of an early WCET using different machine learning approaches. In International Conference on P2P, Parallel, Grid, Cloud and Internet Computing (3PGCIC), 2022. URL: https://doi.org/10.1007/978-3-031-19945-5_30.
  18. Yau-Tsun Steven Li and Sharad Malik. Performance analysis of embedded software using implicit path enumeration. In SIGPLAN workshop on Languages, compilers, & tools for real-time systems, 1995. URL: https://doi.org/10.1145/216636.216666.
  19. Scott Lundberg and Su-In Lee. A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 2017. Google Scholar
  20. Christoph Molnar, Giuseppe Casalicchio, and Bernd Bischl. Interpretable machine learning-a brief history, state-of-the-art and challenges. In ECML PKDD 2020 Workshops, 2021. URL: https://doi.org/10.1007/978-3-030-65965-3_28.
  21. Hugo Reymond, Jean-Luc Béchennec, Mikaël Briday, Sébastien Faucou, Isabelle Puaut, and Erven Rohou. SCHEMATIC: compile-time checkpoint placement and memory allocation for intermittent systems. In IEEE/ACM International Symposium on Code Generation and Optimization, CGO 2024, 2024. URL: https://doi.org/10.1109/CGO57630.2024.10444789.
  22. Hugo Reymond, Hector Chabot, Abderaouf Nassim Amalou, and Isabelle Puaut. MSP430FR5969 basic block worst case energy consumption (WCEC) and worst case execution time (WCET) dataset, April 2024. URL: https://doi.org/10.5281/zenodo.11066623.
  23. Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. " why should i trust you?" explaining the predictions of any classifier. In SIGKDD international conference on knowledge discovery and data mining, 2016. URL: https://doi.org/10.1145/2939672.2939778.
  24. Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. Anchors: High-precision model-agnostic explanations. In Proceedings of the AAAI conference on artificial intelligence, 2018. URL: https://doi.org/10.1609/aaai.v32i1.11491.
  25. David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. Learning representations by back-propagating errors. Nature, 1986. URL: https://doi.org/10.1038/323533a0.
  26. Robert E. Schapire. The boosting approach to machine learning: An overview. Nonlinear estimation and classification, 2003. URL: https://doi.org/10.1007/978-0-387-21579-2_9.
  27. Texas Instruments. MSP430FR5969 user’s guide. URL: https://www.ti.com/lit/ug/slau367p/slau367p.pdf.
  28. MSP430FR5969 product information. URL: https://www.ti.com/product/MSP430FR5969.
  29. Msp430 family instruction set summary. URL: https://www.ti.com/sc/docs/products/micro/msp430/userguid/as_5.pdf.
  30. Simon Wegener, Kris K. Nikov, Jose Nunez-Yanez, and Kerstin Eder. EnergyAnalyzer: Using Static WCET Analysis Techniques to Estimate the Energy Consumption of Embedded Applications. In 21th International Workshop on Worst-Case Execution Time Analysis (WCET 2023), 2023. URL: https://doi.org/10.4230/OASIcs.WCET.2023.9.
  31. Bahram Yarahmadi and Erven Rohou. Compiler Optimizations for Safe Insertion of Checkpoints in Intermittently Powered Systems. In Embedded Computer Systems: Architectures, Modeling, and Simulation, 2020. URL: https://doi.org/10.1007/978-3-030-60939-9_12.
Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail