Avaliando eficiência energética em padrões de algoritmos para computação científica e de alto desempenho
Abstract
The development of new algorithms usually focuses on performance improvement, with little attention given to environmental impact and energy cost in terms of their execution. However, this topic has been gaining greater attention recently. This work aims to objectively demonstrate the energy consumption of programming patterns commonly found in high-performance scientific programs. The measurement of energy consumption was performed using software through the RAPL interface, and other performance metrics will utilize the quantity of executed operations and elapsed time. Tests are conducted by varying the number of threads used, compilation options, amount of memory utilized, and the processors used, aiming to identify the impacts these changes have on energy efficiency. The results demonstrate that energy efficiency is directly affected by the scalability of the algorithm, and that more aggressive compilation optimizations generally increase it.References
Anwar, M., Qadri, S., and Sattar, A. (2013). Green computing and energy consumption issues in the modern age. IOSR Journal of Computer Engineering, 12(6):91–98.
Brownlee, A. E., Adair, J., Haraldsson, S. O., and Jabbo, J. (2021). Exploring the accuracy – energy trade-off in machine learning. In 2021 IEEE/ACM International Workshop on Genetic Improvement (GI), pages 11–18.
Czarnul, P., Proficz, J., and Krzywaniak, A. (2019). Energy-aware high-performance computing: Survey of state-of-the-art tools, techniques, and environments. Scientific Programming, 2019:1–19.
D’Agostino, D., Merelli, I., Aldinucci, M., and Cesini, D. (2021). Hardware and software solutions for energy-efficient computing in scientific programming. Scientific Programming, 2021:1–9.
Hähnel, M., Döbel, B., Völp, M., and Härtig, H. (2012). Measuring energy consumption for short code paths using rapl. SIGMETRICS Perform. Eval. Rev., 40(3):13–17.
Malony, A. D. (2014). Stencil pattern. Disponível: [link]. Acesso: 04/01/2022.
Murugesan, S. and Gangadharan, G. R. (2012). Harnessing green it: Principles and practices. John Wiley & Sons, Inc.
Pandruvada, S. (2014). Running average power limit. Disponível: [link]. Acesso: 06/01/2022.
Paniego, J. M., Gallo, S., Pi Puig, M., Chichizola, F., De Giusti, L., and Balladini, J. (2018). Analysis of rapl energy prediction accuracy in a matrix multiplication application on shared memory. In De Giusti, A. E., editor, Computer Science – CACIC 2017, pages 37–46, Cham. Springer International Publishing.
Rotem, E., Naveh, A., Ananthakrishnan, A., Weissmann, E., and Rajwan, D. (2012). Power-management architecture of the intel microarchitecture code-named sandy bridge. IEEE Micro, 32(2):20–27.
Salama, M. (2020). Green computing, a contribution to save the environment. Disponível: [link]. Acesso: 06/01/2022.
Venkatesh, A., Kandalla, K., and Panda, D. K. (2013). Evaluation of energy characteristics of mpi communication primitives with rapl. In 2013 IEEE International Symposium on Parallel Distributed Processing, Workshops and Phd Forum, pages 938–945.
Voss, M., Asenjo, R., and Reinders, J. (2019). Mapping Parallel Patterns to TBB, pages 233–248. Apress, Berkeley, CA.
Brownlee, A. E., Adair, J., Haraldsson, S. O., and Jabbo, J. (2021). Exploring the accuracy – energy trade-off in machine learning. In 2021 IEEE/ACM International Workshop on Genetic Improvement (GI), pages 11–18.
Czarnul, P., Proficz, J., and Krzywaniak, A. (2019). Energy-aware high-performance computing: Survey of state-of-the-art tools, techniques, and environments. Scientific Programming, 2019:1–19.
D’Agostino, D., Merelli, I., Aldinucci, M., and Cesini, D. (2021). Hardware and software solutions for energy-efficient computing in scientific programming. Scientific Programming, 2021:1–9.
Hähnel, M., Döbel, B., Völp, M., and Härtig, H. (2012). Measuring energy consumption for short code paths using rapl. SIGMETRICS Perform. Eval. Rev., 40(3):13–17.
Malony, A. D. (2014). Stencil pattern. Disponível: [link]. Acesso: 04/01/2022.
Murugesan, S. and Gangadharan, G. R. (2012). Harnessing green it: Principles and practices. John Wiley & Sons, Inc.
Pandruvada, S. (2014). Running average power limit. Disponível: [link]. Acesso: 06/01/2022.
Paniego, J. M., Gallo, S., Pi Puig, M., Chichizola, F., De Giusti, L., and Balladini, J. (2018). Analysis of rapl energy prediction accuracy in a matrix multiplication application on shared memory. In De Giusti, A. E., editor, Computer Science – CACIC 2017, pages 37–46, Cham. Springer International Publishing.
Rotem, E., Naveh, A., Ananthakrishnan, A., Weissmann, E., and Rajwan, D. (2012). Power-management architecture of the intel microarchitecture code-named sandy bridge. IEEE Micro, 32(2):20–27.
Salama, M. (2020). Green computing, a contribution to save the environment. Disponível: [link]. Acesso: 06/01/2022.
Venkatesh, A., Kandalla, K., and Panda, D. K. (2013). Evaluation of energy characteristics of mpi communication primitives with rapl. In 2013 IEEE International Symposium on Parallel Distributed Processing, Workshops and Phd Forum, pages 938–945.
Voss, M., Asenjo, R., and Reinders, J. (2019). Mapping Parallel Patterns to TBB, pages 233–248. Apress, Berkeley, CA.
Published
2023-10-17
How to Cite
ANJOS, Paulo N. M. dos; FAZENDA, Alvaro L..
Avaliando eficiência energética em padrões de algoritmos para computação científica e de alto desempenho. In: UNDERGRADUATE RESEARCH WORKSHOP - SYMPOSIUM ON HIGH PERFORMANCE COMPUTING SYSTEMS (SSCAD), 24. , 2023, Porto Alegre/RS.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2023
.
p. 49-56.
DOI: https://doi.org/10.5753/wscad_estendido.2023.235760.
