Abstract
This paper presents refinements to the execution-cache-memory performance model and a previously published power model for multicore processors. The combination of both enables a very accurate prediction of performance and energy consumption of contemporary multicore processors as a function of relevant parameters such as number of active cores as well as core and Uncore frequencies. Model validation is performed on Intel Sandy Bridge-EP, Broadwell-EP, and AMD Epyc processors. Production-related variations in chip quality are demonstrated through a statistical analysis of the fit parameters obtained on one hundred Broadwell-EP CPUs of the same model. Insights from the models are used to explain the performance- and energy-related behavior of the processors for scalable as well as saturating (i.e., memory-bound) codes. In the process we demonstrate the models’ capability to identify optimal operating points with respect to highest performance, lowest energy-to-solution, and lowest energy-delay product and identify a set of best practices for energy-efficient execution.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Consider, e.g., the 18-core Broadwell-EP chip, which offers 17 different Uncore and 12 different CPU core frequencies, for which a total of 3672 measurements (each with a non-negligible runtime to reach operating temperature equilibrium) are required. In contrast, setting up the model requires only four, six, and nine measurements on the AMD Epyc, Intel Sandy Bridge-EP, and Broadwell-EP processors, respectively.
- 2.
The term Uncore refers to all parts of the chip that are not part of the core design, such as, e.g., shared last-level cache, ring interconnect, and memory controllers.
- 3.
- 4.
The coefficient of variation is used to measure the relative variance of a sample. It is defined as the ratio of the standard deviation \(\sigma \) to the mean \(\mu \) of a sample.
- 5.
For n active cores, the probability of a core’s memory access encountering a busy bus is \(u(n-1)\); when the bus is busy, the penalty \(p_\mathrm {0}\), which increases with the number of cores, is applied.
- 6.
On Sandy and Ivy Bridge processors the Uncore is clocked at the same frequency as the CPU cores and can thus only be set indirectly.
- 7.
Wall clock time can also be used, which essentially mirrors the plot about the y axis.
- 8.
References
Intel 64 and IA-32 Architectures Optimization Reference Manual. Intel Press, June 2016. http://www.intel.com/content/dam/www/public/us/en/documents/manuals/64-ia-32-architectures-optimization-manual.pdf
De Vogeleer, K., Memmi, G., Jouvelot, P., Coelho, F.: The energy/frequency convexity rule: modeling and experimental validation on mobile devices. In: Wyrzykowski, R., Dongarra, J., Karczewski, K., Waśniewski, J. (eds.) PPAM 2013. LNCS, vol. 8384, pp. 793–803. Springer, Heidelberg (2014). https://doi.org/10.1007/978-3-642-55224-3_74
Freeh, V.W., Lowenthal, D.K., Pan, F., Kappiah, N., Springer, R., Rountree, B.L., Femal, M.E.: Analyzing the energy-time trade-off in high-performance computing applications. IEEE Trans. Parallel Distrib. Syst. 18(6), 835–848 (2007). https://doi.org/10.1109/TPDS.2007.1026
Hackenberg, D., Schöne, R., Ilsche, T., Molka, D., Schuchart, J., Geyer, R.: An energy efficiency feature survey of the Intel Haswell processor. In: 2015 IEEE International Parallel and Distributed Processing Symposium Workshop, pp. 896–904, May 2015
Hager, G., Treibig, J., Habich, J., Wellein, G.: Exploring performance and power properties of modern multicore chips via simple machine models. Concurr. Comput. Pract. Exper. (2013). https://doi.org/10.1002/cpe.3180
Hammer, J., Eitzinger, J., Hager, G., Wellein, G.: Kerncraft: a tool for analytic performance modeling of loop kernels. In: Niethammer, C., Gracia, J., Hilbrich, T., Knüpfer, A., Resch, M.M., Nagel, W.E. (eds.) Tools for High Performance Computing 2016, pp. 1–22. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-56702-0_1
Hofmann, J., Hager, G., Wellein, G., Fey, D.: An analysis of core- and chip-level architectural features in four generations of intel server processors. In: Kunkel, J.M., Yokota, R., Balaji, P., Keyes, D. (eds.) ISC 2017. LNCS, vol. 10266, pp. 294–314. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-58667-0_16
Inadomi, Y., Patki, T., Inoue, K., Aoyagi, M., Rountree, B., Schulz, M., Lowenthal, D., Wada, Y., Fukazawa, K., Ueda, M., Kondo, M., Miyoshi, I.: Analyzing and mitigating the impact of manufacturing variability in power-constrained supercomputing. In: Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015, pp. 78:1–78:12. ACM, New York (2015). http://doi.acm.org/10.1145/2807591.2807638
Khabi, D., Küster, U.: Power consumption of kernel operations. In: Resch, M.M., Bez, W., Focht, E., Kobayashi, H., Kovalenko, Y. (eds.) Sustained Simulation Performance 2013, pp. 27–45. Springer, Cham (2013). https://doi.org/10.1007/978-3-319-01439-5_3
Rauber, T., Rünger, G.: Towards an energy model for modular parallel scientific applications. In: 2012 IEEE International Conference on Green Computing and Communications, pp. 523–532, November 2012
Rauber, T., Rünger, G., Schwind, M., Xu, H., Melzner, S.: Energy measurement, modeling, and prediction for processors with frequency scaling. J. Supercomput. 70(3), 1451–1476 (2014). https://doi.org/10.1007/s11227-014-1236-4
Song, S., Su, C., Rountree, B., Cameron, K.W.: A simplified and accurate model of power-performance efficiency on emergent GPU architectures. In: 2013 IEEE 27th International Symposium on Parallel and Distributed Processing, pp. 673–686, May 2013
Stengel, H., Treibig, J., Hager, G., Wellein, G.: Quantifying performance bottlenecks of stencil computations using the Execution-Cache-Memory model. In: Proceedings of the 29th ACM International Conference on Supercomputing, ICS 2015. ACM, New York (2015). http://doi.acm.org/10.1145/2751205.2751240
Wilde, T., Auweter, A., Shoukourian, H., Bode, A.: Taking advantage of node power variation in homogenous HPC systems to save energy. In: Kunkel, J.M., Ludwig, T. (eds.) ISC High Performance 2015. LNCS, vol. 9137, pp. 376–393. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-20119-1_27
Williams, S., Waterman, A., Patterson, D.: Roofline: An insightful visual performance model for multicore architectures. Commun. ACM 52(4), 65–76 (2009). http://doi.acm.org/10.1145/1498765.1498785
Wittmann, M., Hager, G., Zeiser, T., Treibig, J., Wellein, G.: Chip-level and multi-node analysis of energy-optimized lattice Boltzmann CFD simulations. Concurr. Comput. Pract. Exper. 28(7), 2295–2315 (2016). https://doi.org/10.1002/cpe.3489
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Hofmann, J., Hager, G., Fey, D. (2018). On the Accuracy and Usefulness of Analytic Energy Models for Contemporary Multicore Processors. In: Yokota, R., Weiland, M., Keyes, D., Trinitis, C. (eds) High Performance Computing. ISC High Performance 2018. Lecture Notes in Computer Science(), vol 10876. Springer, Cham. https://doi.org/10.1007/978-3-319-92040-5_2
Download citation
DOI: https://doi.org/10.1007/978-3-319-92040-5_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-92039-9
Online ISBN: 978-3-319-92040-5
eBook Packages: Computer ScienceComputer Science (R0)