Estimate energy consumption of web and cloud-based software with models
This note summarizes how to estimate the energy consumption by using software models. If you want to estimate the carbon emissions see note Estimate carbon emissions of cloud applications.
Teads' AWS Estimation Model
Famous model for estimating AWS instances energy usage. This model creates a power curve on a correlation to SPEC Power database. This allows the model to generate a power curve for any AWS EC2 instance type based on publicly available AWS EC2 Instance CPU data.
Original article: Building an AWS EC2 Carbon Emissions Dataset | by Benjamin DAVY | Teads Engineering | Medium
Impact Framework Model: teads-aws
Cloud Energy by Green Coding Solutions
Website: https://www.green-coding.io/projects/cloud-energy/
Source code: https://github.com/green-coding-berlin/spec-power-model
Since in the cloud it is often not possible to measure energy directly we have created a Machine Learning estimation model based on the data from SPECPower
Input data:
- server utilization
- many hardware specs
Impact Framework Model: -
Boavizta
Boavizta is an environmental impact calculator that exposes an API that can be used to retrieve energy and embodied carbon estimates.
Impact Framework Model: boavizta
Methodology: Boavizta#Methodology
Kepler
The Kepler stack comprises Kepler and Kepler Model Server.
The Model Server is used to train power models, and it can be optionally deployed alongside Kepler to help Kepler select the most appropriate power model for a given environment. For example, considering the CPU model, available metrics and the required model accuracy.
The Model Server trains its models using Prometheus metrics from a specific bare-metal node. It records how much energy the node consumed and the resource utilization of containers and system processes (OS and other background processes).
Impact Framework Model: -
Research paper:
🔗 References
Conscientious Computing - Accurately measuring the energy consumption of hardware
Studis
Pathania, P., Mehra, R., Sharma, V. S., Kaulgud, V., Podder, S., & Burden, A. P. (2022). ESAVE: Estimating Server and Virtual Machine Energy. Proceedings of the 37th IEEE/ACM International Conference on Automated Software Engineering, 1–3. https://doi.org/10.1145/3551349.3561170
Rteil, N., Bashroush, R., Kenny, R., & Wynne, A. (2022). Interact: IT infrastructure energy and cost analyzer tool for data centers. Sustainable Computing: Informatics and Systems, 33, 100618. https://doi.org/10.1016/j.suscom.2021.100618
Choochotkaew, S., Wang, C., Chen, H., Chiba, T., Amaral, M., Lee, E. K., & Eilam, T. (2023). Advancing Cloud Sustainability: A Versatile Framework for Container Power Model Training. 2023 31st International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS), 1–4. https://doi.org/10.1109/MASCOTS59514.2023.10387542