Abstract
Amazon web services provide resources equipped with graphics processing unit (GPU) devices, which are commonly used for parallel processing applications. When there is less demand for cloud computing instances, the unused resources are provided at low-cost pricing as GPU spot instances. The pricing of the GPU spot instances dynamically changes over time based on the long-term demand and supply of cloud resources in the spot market. The main aim of this research is to predict upcoming GPU spot instance pricing using a time series prediction model. Our second aim is to study the unique characteristics of GPU spot instances. This research will help researchers to exploit GPU spot instances across distinct regions in a cost-efficient way. In the existing literature, researchers used machine learning approaches including artificial neural networks (ANN), long short-term memory (LSTM) networks, and random forecasts. They also used classical univariate time series approaches including autoregressive integrated moving average (ARIMA) and autoregressive moving average (ARMA) models. However, deep-learning price-reduction methods are complex, require high computational power, and they cannot provide accurate price forecasts due to the varying and volatile environment. As such, the use of time series prediction models is more suitable under these circumstances. But choosing a time series prediction model with different environments including variance and volatility is a challenging task, and none of the extant literature provides such an exploratory analysis. This research used the linear autoregressive (AR) model, ARIMA model, exponential smoothing (ETS) model, and generalized autoregressive conditional heteroskedasticity (GARCH) for predicting upcoming GPU spot instance pricing. This research provides a comparative exploration regarding the use of these prediction models across distinct regions in a cost-efficient way. The results comparisons between AR, ETS, ARIMA, and GARCH models show clearly that the GARCH model provides better results for dynamic based short-term and middle-term GPU instance price prediction. The AR model provides best results for variance based middle-term and long-term GPU instance price prediction. The ETS model provides better results for smoothness based long-term GPU spot price prediction. The ARIMA model gives best results for fluctuations based long-term GPU spot price prediction.
Similar content being viewed by others
Data availability
Data availability is not applicable. No new data is generated.
References
Siavashi, A., Momtazpour, M.: Gpucloudsim: an extension of cloudsim for modeling and simulationof gpus in cloud data centers. J. Supercomput. 75(5), 2535–2561 (2019)
Risco, S., Moltó, G.: GPU-enabled serverless workflows for efficient multimedia processing. Appl. Sci. 11(4), 1438 (2021)
Younge, A.J., Walters, J.P., Crago, S., Fox, G.C.: Evaluating GPU passthrough in Xen for high performance cloud computing. In: 2014 IEEE international parallel & distributed processing symposium workshops, pp. 852–859. IEEE, Piscataway (2014)
Lee, K., Son, M.: Deepspotcloud: leveraging cross-region GPU spot instances for deep learning. In: 2017 IEEE 10th international conference on cloud computing (CLOUD). IEEE, Piscataway (2017)
Lancon, J., Kunwar, Y., Stroud, D., McGee, M., Slater, R.: AWS EC2 instance spot price forecasting using LSTM networks. SMU Data Sci. Rev. 2(2), 8 (2019)
Kumar, D., et al.: A survey on spot pricing in cloud computing. J. Netw. Syst. Manage. 26(4), 809–856 (2018)
Idrees, S.M., Alam, M.A., Agarwal, P.: A prediction approach for stock market volatility based on time series data. IEEE Access 7, 17287–17298 (2019)
Imai, C., Armstrong, B., Chalabi, Z., Mangtani, P., Hashizume, M.: Time series regression model for infectious disease and weather. Environ. Res. 142, 319–327 (2015)
Frees, E.W.: Analytics of insurance markets. Annu. Rev. Financ. Econ. 7, 253–277 (2015)
He, Z.Y., Jin, L.W.: Activity recognition from acceleration data using AR model representation and SVM. In: 2008 international conference on machine learning and cybernetics, vol. 4, pp. 2245–2250. IEEE, Piscataway (2008)
Xu, W., Peng, H., Zeng, X., Zhou, F., Tian, X., Peng, X.: Deep belief network-based AR model for nonlinear time series forecasting. Appl. Soft Comput. 77, 605–621 (2019)
Jain, G., Mallick, B.: A study of time series models ARIMA and ETS. SSRN Electron. J. (2017). https://doi.org/10.2139/ssrn.2898968
Matyjaszek, M., Fernández, P.R., Krzemień, A., Wodarski, K., Valverde, G.F.: Forecasting coking coal prices by means of ARIMA models and neural networks, considering the transgenic time series theory. Resour. Policy 61, 283–292 (2019)
Sánchez, A.B., Ordóñez, C., Lasheras, F.S., de Cos Juez, F.J., Roca-Pardiñas, J.: Forecasting SO2 pollution incidents by means of Elman artificial neural networks and ARIMA models. Abstr. Appl. Anal. (2013). https://doi.org/10.1155/2013/238259
Hu, Y., Ni, J., Wen, L.: A hybrid deep learning approach by integrating LSTM-ANN networks with GARCH model for copper price volatility prediction. Physica A 557, 124907 (2020)
Cheikh, N.B., Zaied, Y.B., Chevallier, J.: Asymmetric volatility in cryptocurrency markets: new evidence from smooth transition GARCH models. Finance Res. Lett. 35, 101293 (2020)
Francq, C., Zakoian, J.: GARCH models structure, statistical inference and financial applications. Wiley, Hoboken (2011)
Zhao, H., et al.: Optimal resource rental planning for elastic applications in cloud market. In: 2012 IEEE 26th international parallel and distributed processing symposium. IEEE, Piscataway (2012)
Agmon Ben-Yehuda, O., et al.: Deconstructing Amazon EC2 spot instance pricing. ACM Trans. Econ. Comput. (TEAC) 1(3), 1–20 (2013)
Khandelwal, V., Chaturvedi, A.K., Gupta, C.P.: Amazon EC2 spot price prediction using regression random forests. IEEE Trans. Cloud Comput. 8(1), 59–72 (2020). https://doi.org/10.1109/TCC.2017.2780159
Wallace, R.M., et al.: Applications of neural-based spot market prediction for cloud computing. In: 2013 IEEE 7th international conference on intelligent data acquisition and advanced computing systems (IDAACS), pp. 710–716. IEEE, Piscataway (2013). https://doi.org/10.1109/IDAACS.2013.6663017
Baughman, M., Haas, C., Wolski, R., Foster, I., Chard, K.: Predicting amazon spot prices with LSTM networks. In: Proceedings of the 9th workshop on scientific cloud computing 2018—co-located with HPDC. (2018)
Arevalos, S., Lopez-Pires, F., Baran, B.: A comparative evaluation of algorithms for auction-based cloud pricing prediction. In: Proc.−2016 IEEE Int. Conf. Cloud Eng. IC2E 2016 co-located with 1st IEEE Int. Conf. Internet-of-Things Des. Implementation, IoTDI, pp. 99–108 (2016)
Portella, G., et al.: Statistical analysis of Amazon EC2 cloud pricing models. Concurr Comput: Pract Exp 31(18), 4451 (2019)
Alkharif, S., Lee, K., Kim, H.: Time-series analysis for price prediction of opportunistic cloud computing resources. In: Proceedings of the 7th international conference on emerging databases. Springer, Singapore (2018)
Cai, Z., Li, X., Ruiz, R., Li, Q.: Price forecasting for spot instances in cloud computing. Future Gener. Comput. Syst. 79, 38–53 (2018)
Neto, J.P.A., Pianto, D.M., Ralha, C.G.: A prediction approach to define checkpoint intervals in spot instances. In: Luo, M., Zhang, L.J. (eds.) Cloud computing—CLOUD 2018 lecture notes in computer science, vol. 10967, pp. 84–93. Springer, Cham (2018)
Mishra, A.K., Yadav, D.K.: Analysis and prediction of Amazon EC2 spot instance prices. Int. J. Appl. Eng. Res. 21, 11205–11212 (2017)
Kong, D., Liu, S., Pan, L.: Amazon spot instance price prediction with GRU network. In: 2021 IEEE 24th international conference on computer supported cooperative work in design (CSCWD), pp. 31–36. IEEE, Piscataway (2021)
Chittora, V., Gupta, C.P.: Dynamic spot price forecasting using stacked LSTM networks. In: 2020 3rd international conference on intelligent sustainable systems (ICISS), pp. 1080–1085. IEEE. (2020)
Isobe, G., Aketani, S., Koita, T.: An evaluation of the discount rates for spot instances on Amazon EC2. Bull. Netw. Comput. Syst. Softw. 10(1), 27–29 (2021)
Arevalos, S., Lopez-Pires, F., Baran, B.: A comparative evaluation of algorithms for auction-based cloud pricing prediction. In: Proc.—2016 IEEE Int. Conf. Cloud Eng. IC2E 2016 co-located with 1st IEEE Int. Conf. Internet-of-Things Des. Implementation, IoTDI 2016, pp. 99–108 (2016)
Abualigah, L., Diabat, A., Sumari, P., Gandomi, A.H.: A novel evolutionary arithmetic optimization algorithm for multilevel thresholding segmentation of covid-19 ct images. Processes 9(7), 1155 (2021)
Abualigah, L., Abd Elaziz, M., Sumari, P., Geem, Z.W., Gandomi, A.H.: Reptile search algorithm (RSA): a nature-inspired meta-heuristic optimizer. Expert Syst. Appl. 191, 116158 (2022)
Abualigah, L., Yousri, D., Abd Elaziz, M., Ewees, A.A., Al-qaness, M.A., Gandomi, A.H.: Aquila Optimizer: a novel meta-heuristic optimization algorithm. Comput. Ind. Eng. 157, 107250 (2021)
Abualigah, L., Diabat, A., Abd Elaziz, M.: Intelligent workflow scheduling for big data applications in IoT cloud computing environments. Clust. Comput. (2021). https://doi.org/10.1007/s10586-021-03291-7
Javadi, B., Thulasiramy, R.K., Buyya, R.: Statistical modeling of spot instance prices in public cloud environments. In: 2011 fourth IEEE international conference on utility and cloud computing, pp. 219–228. IEEE, Piscataway (2011)
Agarwal, S., Mishra, A.K., Yadav, D.K.: Forecasting price of amazon spot instances using neural networks. Int. J. Appl. Eng. Res 12(20), 10276–10283 (2017)
AWS Command Line Interface. https://aws.amazon.com/cli/, Accessed 2 Mar 2019
AWS, Y.M.U., Singh, H.: Practical machine learning with AWS
Liu, W., Wang, P., Meng, Y., Zhao, C., Zhang, Z.: Cloud spot instance price prediction using k NN regression. HCIS 10(1), 1–14 (2020)
Baron, J., Kotecha, S.: Storage options in the aws cloud. Amazon Web Services, Washington DC, Tech. Rep (2013)
Tan, Z., Zhang, J., Wang, J., Xu, J.: Day-ahead electricity price forecasting using wavelet transform combined with ARIMA and GARCH models. Appl. Energy 87(11), 3606–3610 (2010)
Murali, P., Revathy, R., Balamurali, S., Tayade, A.S.: Integration of RNN with GARCH refined by whale optimization algorithm for yield forecasting: a hybrid machine learning approach. J. Ambient Intell. Humaniz. Comput. (2020). https://doi.org/10.1007/s12652-020-01922-2
Funding
There is no funding source available for this research work.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
The authors have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Khan, M., Jehangiri, A.I., Ahmad, Z. et al. An exploration to graphics processing unit spot price prediction. Cluster Comput 25, 3499–3515 (2022). https://doi.org/10.1007/s10586-022-03581-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10586-022-03581-8