Search

From HPC to Green Computing: Quantifying and Reducing Environmental Impact through Mathematical Models

Author(s)

Anqi Sun1,*,#, Mengfei Song1,#, Ao Xiong2,#, Xiuqi Zhu3,#

Affiliation(s)

1Shenzhen Foreign Languages School, Shenzhen, Guangdong, China 2Beijing Etown Academy, Beijing, China 3High School Affiliated to Shanghai Jiao Tong University, Shanghai, China *Corresponding Author. #These authors contribute equally to this work.

Abstract

To explore optimized high-performance computing energy consumption and environmental impact, we applied different mathematical models to solve green computing problems in distinct scenarios. First of all, we quantified global HPC energy consumption at full load by analyzing total energy consumption and power consumption across years and countries. Next, we developed a comprehensive model to estimate annual global HPC system carbon emissions based on HPC power consumption, energy mix data, and emission factors for various energy sources. Then to estimate future HPC carbon emissions, we fitted an effective Elastic Net Regression Model combining L1 and L2 regularization based on energy mix proportions, HPC energy use, and carbon emissions data from 2014–2023. The results show a peak in annual emissions in 2017 (1.456 → 1010 kg CO2), followed by a decline to 6.99 → 109 kg in 2023, indicating a shift towards renewable energy sources. In addition, using the Game Theoretical Model, we analyze competition and cooperation among HPC data centers in water resource allocation. Finally, we objectively analyzed the strengths and weaknesses of the above-mentioned model. We also drafted a non-technical report for the United Nations Advisory Board letter, using the results of our assessment and taking environmental effects into consideration.

Keywords

HPC; Elastic Network Regression; Environmental Protection; Dynamic System Modeling; Game Theory Model

References

[1] Hasan Mostafaei, Hadi Bahmani, Davood Mostofinejad, and Chengqing Wu. A novel development of hpc without cement: Mechanical properties and sustainability evaluation. Journal of Building Engineering, 76:107262, 2023. [2] Baolin Li, Rohan Basu Roy, Daniel Wang, Siddharth Samsi, Vijay Gadepally, and Devesh Tiwari. Toward sustainable hpc: Carbon footprint estimation and environmental implications of hpc systems. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, pages 1––15, 2023. [3] William Fox, Devarshi Ghoshal, Abel Souza, Gonzalo P Rodrigo, and Lavanya Ramakrishnan. E-hpc: a library for elastic resource management in hpc environments. In Proceedings of the 12th Workshop on Workflows in Support of Large-Scale Science, pages 1–11, 2017. [4] Kenneth O’brien, Ilia Pietri, Ravi Reddy, Alexey Lastovetsky, and Rizos Sakellariou. A survey of power and energy predictive models in hpc systems and applications. ACM Computing Surveys (CSUR), 50(3):1–38, 2017. [5] Jingwei Sun, Guangzhong Sun, Shiyan Zhan, Jiepeng Zhang, and Yong Chen. Automated performance modeling of hpc applications using machine learning. IEEE Transactions on Computers, 69(5):749–763, 2020. [6] William F. C. Tavares, Marcio Roberto Miranda Assis, and Edson Borin. Quantifying and detecting hpc resource wastage in cloud environments. In 2021 International Symposium on Computer Architecture and High Performance Computing Workshops (SBAC-PADW), pages 41– 46, 2021. [7] Eduardo Camilo Inacio and Mario AR Dantas. A survey into performance and energy efficiency in hpc, cloud and big data environments. International Journal of Networking and Virtual Organisations, 14(4):299–318, 2014. [8] Shuangshuang Jin, Zhenyu Huang, Ruisheng Diao, Di Wu, and Yousu Chen. Comparative implementation of highperformance computing for power system dynamic simulations. IEEE Transactions on Smart Grid, 8(3):1387–1395, 2017. [9] Cheng, Zheng Yuan. "Modeling cooperative and competitive behaviors in emergency evacuation: A game-theoretical approach." Computers & Mathematics with Applications (2011). [10]Ishfaq Ahmad, Sanjay Ranka, and Samee Ullah Khan. Using game theory for scheduling tasks on multi-core processors for simultaneous optimization of performance and energy. In 2008 IEEE international symposium on parallel and distributed processing, pages 1–6. IEEE, 2008. [11]Liang, Quincy. "AMOLED market revenue to grow by about 10 times 2010-2015: Display Bank. (activematrix organic light-emit diode) (Brief article)." [12]Graymore, Michelle L. M., and A. M. Wallis. "Water savings or water efficiency? Water-use attitudes and behaviour in rural and regional areas." International Journal of Sustainable Development & World Ecology 17.1(2010):84-93.