IDEAS home Printed from https://ideas.repec.org/a/gam/jeners/v16y2023i5p2255-d1081268.html
   My bibliography  Save this article

Modular and Transferable Machine Learning for Heat Management and Reuse in Edge Data Centers

Author

Listed:
  • Rickard Brännvall

    (ICE Data Center, RISE Research Institutes of Sweden AB, 973 47 Luleå, Sweden
    EISLAB, Luleå University of Technology, 971 87 Luleå, Sweden)

  • Jonas Gustafsson

    (ICE Data Center, RISE Research Institutes of Sweden AB, 973 47 Luleå, Sweden)

  • Fredrik Sandin

    (EISLAB, Luleå University of Technology, 971 87 Luleå, Sweden)

Abstract

This study investigates the use of transfer learning and modular design for adapting a pretrained model to optimize energy efficiency and heat reuse in edge data centers while meeting local conditions, such as alternative heat management and hardware configurations. A Physics-Informed Data-Driven Recurrent Neural Network (PIDD RNN) is trained on a small scale-model experiment of a six-server data center to control cooling fans and maintain the exhaust chamber temperature within safe limits. The model features a hierarchical regularizing structure that reduces the degrees of freedom by connecting parameters for related modules in the system. With a RMSE value of 1.69, the PIDD RNN outperforms both a conventional RNN (RMSE: 3.18), and a State Space Model (RMSE: 2.66). We investigate how this design facilitates transfer learning when the model is fine-tuned over a few epochs to small dataset from a second set-up with a server located in a wind tunnel. The transferred model outperforms a model trained from scratch over hundreds of epochs.

Suggested Citation

  • Rickard Brännvall & Jonas Gustafsson & Fredrik Sandin, 2023. "Modular and Transferable Machine Learning for Heat Management and Reuse in Edge Data Centers," Energies, MDPI, vol. 16(5), pages 1-24, February.
    • Handle: RePEc:gam:jeners:v:16:y:2023:i:5:p:2255-:d:1081268
    as

    Download full text from publisher

    File URL: https://www.mdpi.com/1996-1073/16/5/2255/pdf
    Download Restriction: no
    File URL: https://www.mdpi.com/1996-1073/16/5/2255/
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Manaserh, Yaman M. & Tradat, Mohammad I. & Bani-Hani, Dana & Alfallah, Aseel & Sammakia, Bahgat G. & Nemati, Kourosh & Seymour, Mark J., 2022. "Machine learning assisted development of IT equipment compact models for data centers energy planning," Applied Energy, Elsevier, vol. 305(C).
    2. Gokhale, Gargya & Claessens, Bert & Develder, Chris, 2022. "Physics informed neural networks for control oriented thermal modeling of buildings," Applied Energy, Elsevier, vol. 314(C).
    3. Geyer, Philipp & Singaravel, Sundaravelpandian, 2018. "Component-based machine learning for performance prediction in building design," Applied Energy, Elsevier, vol. 228(C), pages 1439-1453.
    4. Emelie Wibron & Anna-Lena Ljung & T. Staffan Lundström, 2018. "Computational Fluid Dynamics Modeling and Validating Experiments of Airflow in a Data Center," Energies, MDPI, vol. 11(3), pages 1-15, March.
    5. Habibi Khalaj, Ali & Halgamuge, Saman K., 2017. "A Review on efficient thermal management of air- and liquid-cooled data centers: From chip to the cooling system," Applied Energy, Elsevier, vol. 205(C), pages 1165-1188.
    6. Wang, Xinyue & Liu, Yang & Tian, Tong & Li, Ji, 2022. "Directly air-cooled compact looped heat pipe module for high power servers with extremely low power usage effectiveness," Applied Energy, Elsevier, vol. 319(C).
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Du, Yahui & Zhou, Zhihua & Yang, Xiaochen & Yang, Xueqing & Wang, Cheng & Liu, Junwei & Yuan, Jianjuan, 2023. "Dynamic thermal environment management technologies for data center: A review," Renewable and Sustainable Energy Reviews, Elsevier, vol. 187(C).
    2. Tian, Tong & Wang, Xinyue & Liu, Yang & Yang, Xuan & Sun, Bo & Li, Ji, 2023. "Nano-engineering enabled heat pipe battery: A powerful heat transfer infrastructure with capability of power generation," Applied Energy, Elsevier, vol. 348(C).
    3. Chen, Dong & Chui, Chee-Kong & Lee, Poh Seng, 2025. "Adaptive physically consistent neural networks for data center thermal dynamics modeling," Applied Energy, Elsevier, vol. 377(PD).
    4. Zhang, Qingang & Zeng, Wei & Lin, Qinjie & Chng, Chin-Boon & Chui, Chee-Kong & Lee, Poh-Seng, 2023. "Deep reinforcement learning towards real-world dynamic thermal management of data centers," Applied Energy, Elsevier, vol. 333(C).
    5. Zhou, Haojie & Tian, Tong & Wang, Xinyue & Li, Ji, 2023. "Combining looped heat pipe and thermoelectric generator module to pursue data center servers with possible power usage effectiveness less than 1," Applied Energy, Elsevier, vol. 332(C).
    6. Tian, Tong & Yang, Xuan & Li, Ji, 2025. "To boost waste heat harvesting and power generation through a portable heat pipe battery during high efficient electronics cooling," Applied Energy, Elsevier, vol. 377(PA).
    7. Shen, Yuxuan & Pan, Yue, 2023. "BIM-supported automatic energy performance analysis for green building design using explainable machine learning and multi-objective optimization," Applied Energy, Elsevier, vol. 333(C).
    8. Amal A. Al-Shargabi & Abdulbasit Almhafdy & Dina M. Ibrahim & Manal Alghieth & Francisco Chiclana, 2021. "Tuning Deep Neural Networks for Predicting Energy Consumption in Arid Climate Based on Buildings Characteristics," Sustainability, MDPI, vol. 13(22), pages 1-20, November.
    9. Lu, Tao & Lü, Xiaoshu & Välisuo, Petri & Zhang, Qunli & Clements-Croome, Derek, 2024. "Innovative approaches for deep decarbonization of data centers and building space heating networks: Modeling and comparison of novel waste heat recovery systems for liquid cooling systems," Applied Energy, Elsevier, vol. 357(C).
    10. Liang, Xinbin & Zhu, Xu & Chen, Siliang & Jin, Xinqiao & Xiao, Fu & Du, Zhimin, 2023. "Physics-constrained cooperative learning-based reference models for smart management of chillers considering extrapolation scenarios," Applied Energy, Elsevier, vol. 349(C).
    11. Lahoucine Ouhsaine & Mohammed El Ganaoui & Abdelaziz Mimet & Jean-Michel Nunzi, 2021. "A Substitutive Coefficients Network for the Modelling of Thermal Systems: A Mono-Zone Building Case Study," Energies, MDPI, vol. 14(9), pages 1-19, April.
    12. Di Natale, L. & Svetozarevic, B. & Heer, P. & Jones, C.N., 2023. "Towards scalable physically consistent neural networks: An application to data-driven multi-zone thermal building models," Applied Energy, Elsevier, vol. 340(C).
    13. Ina VIRTOSU & Chen LI, 2024. "Smart life and sustainable development: a comparative analysis on energy and water efficiency in China and the EU," Smart Cities and Regional Development (SCRD) Journal, Smart-EDU Hub, Faculty of Public Administration, National University of Political Studies & Public Administration, vol. 8(3), pages 19-40, April.
    14. Singh, Manav Mahan & Singaravel, Sundaravelpandian & Geyer, Philipp, 2021. "Machine learning for early stage building energy prediction: Increment and enrichment," Applied Energy, Elsevier, vol. 304(C).
    15. Ye, Guisen & Gao, Feng & Fang, Jingyang, 2022. "A mission-driven two-step virtual machine commitment for energy saving of modern data centers through UPS and server coordinated optimizations," Applied Energy, Elsevier, vol. 322(C).
    16. Xiaofei Huang & Junwei Yan & Xuan Zhou & Yixin Wu & Shichen Hu, 2023. "Cooling Technologies for Internet Data Center in China: Principle, Energy Efficiency, and Applications," Energies, MDPI, vol. 16(20), pages 1-31, October.
    17. Zhao, Haitao & Ezeh, Collins I. & Ren, Weijia & Li, Wentao & Pang, Cheng Heng & Zheng, Chenghang & Gao, Xiang & Wu, Tao, 2019. "Integration of machine learning approaches for accelerated discovery of transition-metal dichalcogenides as Hg0 sensing materials," Applied Energy, Elsevier, vol. 254(C).
    18. Abdo Abdullah Ahmed Gassar & Choongwan Koo & Tae Wan Kim & Seung Hyun Cha, 2021. "Performance Optimization Studies on Heating, Cooling and Lighting Energy Systems of Buildings during the Design Stage: A Review," Sustainability, MDPI, vol. 13(17), pages 1-47, September.
    19. Xiao, Tianqi & You, Fengqi, 2023. "Building thermal modeling and model predictive control with physically consistent deep learning for decarbonization and energy optimization," Applied Energy, Elsevier, vol. 342(C).
    20. Zhang, Yingbo & Shan, Kui & Li, Xiuming & Li, Hangxin & Wang, Shengwei, 2023. "Research and Technologies for next-generation high-temperature data centers – State-of-the-arts and future perspectives," Renewable and Sustainable Energy Reviews, Elsevier, vol. 171(C).

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gam:jeners:v:16:y:2023:i:5:p:2255-:d:1081268. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: MDPI Indexing Manager (email available below). General contact details of provider: https://www.mdpi.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.