Comparison of Cloud-Based Free-Tier Virtual Environments: A Performance Analysis on Machine Learning Training
Keywords:
Cloud computing, Machine Learning Classification, Google ColaboratoryAbstract
The increasing demand for machine learning (ML) across various domains has driven the need for accessible training environments, particularly for researchers and students lacking high-end hardware. Cloud-based platforms such as Google Colaboratory and Kaggle Notebooks offer free-tier access to computational resources, making them popular options for model development. However, limited comparative research exists on their real-world performance when training ML models. This study introduces a systematic, experiment-driven benchmarking framework that directly compares these two platforms under identical conditions. The key idea of this approach is to evaluate platform efficiency by training both deep learning (Convolutional Neural Networks) and traditional machine learning (Decision Trees) models on a standardized dataset (CIFAR-10), while capturing quantitative metrics such as training time, memory usage, and CPU utilization. Unlike prior studies, which focus on individual platform capabilities, this work provides a side-by-side, reproducible comparison that reveals how platform design impacts performance for different model types. Results show that Kaggle Notebooks outperform Google Colaboratory, achieving 62% faster training for CNNs and 38% faster for Decision Trees, with lower memory and CPU usage. The findings contribute new insights for students, researchers, and practitioners when choosing cloud-based free-tier platforms for machine learning development.
References
T. Carneiro, R. V. Medeiros Da NóBrega, T. Nepomuceno, G.-B. Bian, V. H. C. De Albuquerque, and P. P. R. Filho, “Performance Analysis of Google Colaboratory as a Tool for Accelerating Deep Learning Applications,” IEEE Access, vol. 6, pp. 61677–61685, 2018, doi: 10.1109/ACCESS.2018.2874767.
A. Ali et al., “Enabling scalable and adaptive machine learning training via serverless computing on public cloud,” Perform. Eval., vol. 167, p. 102451, Mar. 2025, doi: 10.1016/j.peva.2024.102451.
P. Gimeno Sarroca and M. Sánchez-Artigas, “MLLess: Achieving cost efficiency in serverless machine learning training,” J. Parallel Distrib. Comput., vol. 183, p. 104764, Jan. 2024, doi: 10.1016/j.jpdc.2023.104764.
B. Cottier, R. Rahman, L. Fattorini, N. Maslej, T. Besiroglu, and D. Owen, “The rising costs of training frontier AI models,” Feb. 07, 2025, arXiv: arXiv:2405.21015. doi: 10.48550/arXiv.2405.21015.
D. Chahal, M. Mishra, S. Palepu, and R. Singhal, “Performance and Cost Comparison of Cloud Services for Deep Learning Workload,” in Companion of the ACM/SPEC International Conference on Performance Engineering, in ICPE ’21. New York, NY, USA: Association for Computing Machinery, Apr. 2021, pp. 49–55. doi: 10.1145/3447545.3451184.
R. Pakdel and J. Herbert, “Adaptive Cost Efficient Framework for Cloud-Based Machine Learning,” in 2017 IEEE 41st Annual Computer Software and Applications Conference (COMPSAC), Jul. 2017, pp. 155–160. doi: 10.1109/COMPSAC.2017.42.
E. Guerra, F. Wilhelmi, M. Miozzo, and P. Dini, “The Cost of Training Machine Learning Models Over Distributed Data Sources,” IEEE Open J. Commun. Soc., vol. 4, pp. 1111–1126, 2023, doi: 10.1109/OJCOMS.2023.3274394.
R. Teixeira, L. Almeida, M. Antunes, D. Gomes, and R. L. Aguiar, “Efficient training: Federated learning cost analysis,” Big Data Res., vol. 40, p. 100510, May 2025, doi: 10.1016/j.bdr.2025.100510.
S. M. Rajagopal, S. M., and R. Buyya, “Leveraging blockchain and federated learning in Edge-Fog-Cloud computing environments for intelligent decision-making with ECG data in IoT,” J. Netw. Comput. Appl., vol. 233, p. 104037, Jan. 2025, doi: 10.1016/j.jnca.2024.104037.
J. Lawrence, J. Malmsten, A. Rybka, D. A. Sabol, and K. Triplin, “Comparing TensorFlow Deep Learning Performance Using CPUs, GPUs, Local PCs and Cloud”.
P. Osypanka and P. Nawrocki, “Resource Usage Cost Optimization in Cloud Computing Using Machine Learning,” IEEE Trans. Cloud Comput., vol. 10, no. 3, pp. 2079–2089, Jul. 2022, doi: 10.1109/TCC.2020.3015769.
M. T. Islam, S. Karunasekera, and R. Buyya, “Performance and Cost-Efficient Spark Job Scheduling Based on Deep Reinforcement Learning in Cloud Computing Environments,” IEEE Trans. Parallel Distrib. Syst., vol. 33, no. 7, pp. 1695–1710, Jul. 2022, doi: 10.1109/TPDS.2021.3124670.
T. Khan, W. Tian, G. Zhou, S. Ilager, M. Gong, and R. Buyya, “Machine learning (ML)-centric resource management in cloud computing: A review and future directions,” J. Netw. Comput. Appl., vol. 204, p. 103405, Aug. 2022, doi: 10.1016/j.jnca.2022.103405.
G. Kasinathan and S. Jayakumar, “Cloud-Based Lung Tumor Detection and Stage Classification Using Deep Learning Techniques,” BioMed Res. Int., vol. 2022, no. 1, p. 4185835, 2022, doi: 10.1155/2022/4185835.
V. Lahoura et al., “Cloud Computing-Based Framework for Breast Cancer Diagnosis Using Extreme Learning Machine,” Diagnostics, vol. 11, no. 2, Art. no. 2, Feb. 2021, doi: 10.3390/diagnostics11020241.
G. Loukas, T. Vuong, R. Heartfield, G. Sakellari, Y. Yoon, and D. Gan, “Cloud-Based Cyber-Physical Intrusion Detection for Vehicles Using Deep Learning,” IEEE Access, vol. 6, pp. 3491–3508, 2018, doi: 10.1109/ACCESS.2017.2782159.
C. Tong, X. Yin, S. Wang, and Z. Zheng, “A novel deep learning method for aircraft landing speed prediction based on cloud-based sensor data,” Future Gener. Comput. Syst., vol. 88, pp. 552–558, Nov. 2018, doi: 10.1016/j.future.2018.06.023.
D. Wu, C. Jennings, J. Terpenny, and S. Kumara, “Cloud-based machine learning for predictive analytics: Tool wear prediction in milling,” in 2016 IEEE International Conference on Big Data (Big Data), Dec. 2016, pp. 2062–2069. doi: 10.1109/BigData.2016.7840831.
S. Staravoitau, “Efficiency of Machine Learning Cloud-Based Services vs Traditional Methods in Stock Prices Prediction”.
Á. López García et al., “A Cloud-Based Framework for Machine Learning Workloads and Applications,” IEEE Access, vol. 8, pp. 18681–18692, 2020, doi: 10.1109/ACCESS.2020.2964386.
R. Pakdel, “Cloud-based Machine Learning Architecture for Big Data Analysis”.
J. Sekar and A. Llc, “Deep Learning As A Service (Dlaas) In Cloud Computing: Performance And Scalability Analysis,” vol. 10, no. 3, 2023.
G. P. Selvarajan, “Optimising Machine Learning Workflows In Snowflakedb: A Comprehensive Framework Scalable Cloud-Based Data Analytics,” vol. 8, no. 11, 2021.
“Demo 2B: Google Colab for teaching CS and ML (SIGCSE TS 2024 - Demos) - SIGCSE TS 2024,” Sigcse.org, 2024. https://sigcse2024.sigcse.org/details/sigcse-ts-2024-demos/9/Demo-2B-Google-Colab-for-teaching-CS-and-ML (accessed Aug. 02, 2025).
L. Quaranta, F. Calefato, and F. Lanubile, “KGTorrent: A Dataset of Python Jupyter Notebooks from Kaggle,” IEEE Xplore, May 01, 2021. https://ieeexplore.ieee.org/abstract/document/9463068/ (accessed Dec. 13, 2022).
A. Felias, M. A. Remaldora, R. F. Vicente, and M. J. Timogan, “Forensic Facial Sketch Verification Using Image Processing Algorithms,” SMCC Higher Education Research Journal, vol. 3, no. 1, Jun. 2020, doi: https://doi.org/10.18868/ccs.03.060120.06.
M. J. Timogan, “Implementation of Frequent Pattern (FP)-Growth Algorithm for Comparative Analysis on Teacher’s Performance Evaluation Results,” SMCC Higher Education Research Journal, vol. 3, no. 1, Jun. 2020, doi: https://doi.org/10.18868/cp-v3-a3.
