Flexibility Modelling and Optimization for General Computing Task Loads in Data Centers

LUO Aijia; TIAN Nianfeng; CHEN Min; GUO Qinglai

doi:10.7500/AEPS20251117004

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Flexibility Modelling and Optimization for General Computing Task Loads in Data Centers

Computing Power-Electric Power Coordination | 更新时间：2026-10-10

- Flexibility Modelling and Optimization for General Computing Task Loads in Data Centers
- Automation of Electric Power Systems Vol. 50, Issue 19, Pages: 35-43(2026)
- 作者机构：
  
  清华大学电机工程与应用电子技术系，北京市 100084
- 作者简介：
- 基金信息：
  
  Smart-Grid National Science and Technology Major Project(2025ZD0804800)
- DOI：10.7500/AEPS20251117004
  CLC：
- Received：17 November 2025，
  
  Revised：2026-01-25，
  
  Online First：09 April 2026，
  
  Published：10 October 2026
- 稿件说明：
移动端阅览
罗艾佳,田年丰,陈敏等.数据中心通算任务负荷灵活性建模与优化[J].电力系统自动化,2026,50(19):35-43.

LUO Aijia,TIAN Nianfeng,CHEN Min,et al.Flexibility Modelling and Optimization for General Computing Task Loads in Data Centers[J].Automation of Electric Power Systems,2026,50(19):35-43.
罗艾佳,田年丰,陈敏等.数据中心通算任务负荷灵活性建模与优化[J].电力系统自动化,2026,50(19):35-43. DOI： 10.7500/AEPS20251117004.

LUO Aijia,TIAN Nianfeng,CHEN Min,et al.Flexibility Modelling and Optimization for General Computing Task Loads in Data Centers[J].Automation of Electric Power Systems,2026,50(19):35-43. DOI： 10.7500/AEPS20251117004.

摘要

在全球数字化转型与人工智能加速发展的背景下，数据中心已成为电力系统中的重要负荷。面向数据中心中占比高、可中断性较强的通用算力任务，文中提出一种基于价格信号引导、兼顾任务性能和用电成本的通算任务负荷灵活性建模与优化方法。首先，以线程数和处理器频率为关键可调参数，建立单服务器场景下执行任务功率与完成任务时长关联的负荷用能特性模型，分析多服务器下不同任务分配方式对能效的影响，进而给出通算任务的上调、下调功率可调范围及相应时间窗口，实现通算任务负荷灵活性刻画。结合电价信号激励挖掘数据中心负荷灵活性，构建通算任务性能-能效协同优化模型，联合优化任务分配、线程数与处理器频率，规划不同电价时段的数据中心用能水平。在满足任务时限和服务器资源约束的前提下，降低运行成本。算例结果表明，所构建模型能准确刻画通算任务用能特性，灵活性评估验证了其显著的调节潜力；基于电价的优化调度在满足约束的同时重塑负荷并降低成本，为数据中心细粒度参与电力系统灵活调节提供了可行路径。

Abstract

Against the backdrop of global digital transformation and the accelerated development of artificial intelligence

data centers have become significant loads in power systems. Focusing on high-proportion

highly interruptible general computing tasks in data centers

this paper proposes a price-signal-guided modeling and optimization method for the load flexibility of general computing tasks

while balancing task performance and energy costs. Firstly

an energy-consumption characteristic model correlating task execution power with completion time in a single-server scenario is established

using the number of threads and processor frequency as key adjustable parameters. The impact of different task allocation methods on energy efficiency in a multi-server environment is analysed

thereby providing adjustable ranges for increasing and decreasing power for general computing tasks

along with corresponding time windows

to characterise the load flexibility of general computing tasks. By leveraging electricity price signals to incentivise the exploitation of load flexibility in data centers

a collaborative optimization model for the performance and energy efficiency of general computing tasks is constructed. By jointly optimizing task allocation

the number of threads and processor frequency

and planning data center energy consumption levels across different tariff periods

operation costs are reduced whilst meeting task deadlines and server resource constraints. The results of the case studies demonstrate that the developed model accurately captures the energy-consumption characteristics of general computing tasks

while the flexibility assessment confirms its significant potential for load regulation. Optimal dispatch based on electricity prices reshapes the load and reduces costs while meeting constraints. This provides a viable pathway for data centers to participate in the flexible regulation of the power system at a granular level.

关键词

Keywords

references

International Energy Agency . Electricity 2024： analysis and forecast to 2026 ［EB/OL］. ［ 2025-11-17 ］. https：//iea.blob.core.windows.net/assets/18f3ed24-4b26-4c83-a3d2-8a1be51c8cc8/Electricity2024-Analysisandforecastto2026.pdf https://iea.blob.core.windows.net/assets/18f3ed24-4b26-4c83-a3d2-8a1be51c8cc8/Electricity2024-Analysisandforecastto2026.pdf .

中国信息通信研究院 . 数据中心全生命周期绿色算力指数白皮书［R］. 北京：中国信息通信研究院， 2023 .

China Academy of Information and Communications Technology . White paper on green computing power index for full life cycle of data centers ［R］. Beijing ： China Academy of Information and Communications Technology ， 2023 .

郭亮 . 数据中心发展综述［J］. 信息通信技术与政策， 2023 ， 49 （ 5 ）： 2 - 8 .

Guo Liang . Overview of data center development ［J］. Information and Communications Technology and Policy ， 2023 ， 49 （ 5 ）： 2 - 8 .

中国信息通信研究院 . 全球数字经济白皮书（ 2022 年）［R/OL］. 2025-11-17 . http：//www.caict.ac.cn/kxyj/qwfb/bps/202212/P020221207397428021671.pdf http://www.caict.ac.cn/kxyj/qwfb/bps/202212/P020221207397428021671.pdf .

China Academy of Information and Communications Technology （CAICT） . Global digital economy white paper （ 2022 ）［R/OL］. 2025-11-17 . http：//www.caict.ac.cn/kxyj/qwfb/bps/202212/P020221207397428021671.pdf http://www.caict.ac.cn/kxyj/qwfb/bps/202212/P020221207397428021671.pdf .

Dean J ， Barroso L A . The tail at scale ［J］. Communications of the ACM ， 2013 ， 56 （ 2 ）： 74 - 80 .

Verma A ， Pedrosa L ， Korupolu M ， et al . Large-scale cluster management at google with borg ［C］// Proceedings of the Tenth European Conference on Computer Systems ， 2015 .

Mattson P ， Cheng C ， Coleman C ， et al . MLPerf Training Benchmark ［EB/OL］. ［ 2025-11-17 ］. https：//arxiv.org/abs/1910.01500 https://arxiv.org/abs/1910.01500 .

Yoo A B ， Jette M A ， Grondona M . SLURM： simple linux utility for resource management ［C］// The 9th International Workshop ， JSSPP 2003 ， 2003 .

Feitelson D G ， Rudolph L ， Schwiegelshohn U ， et al . Theory and practice in parallel job scheduling ［C］// IPPS’97 Workshop ， 1997 .

Dean J ， Ghemawat S . MapReduce： simplified data processing on large clusters ［J］. Communications of the ACM ， 2008 ， 51 （ 1 ）： 107 - 113 .

Chen S C ， Li Q J ， Zhou M C ， et al . Recent advances in collaborative scheduling of computing tasks in an edge computing paradigm ［J］. Sensors ， 2021 ， 21 （ 3 ）： 779 .

Balsini A ， Pannocchi L ， Cucinotta T . Modeling and simulation of power consumption and execution times for real-time tasks on embedded heterogeneous architectures ［J］. ACM SIGBED Review ， 2019 ， 16 （ 3 ）： 51 - 56 .

Leon-Vega L ， Tosato N ， Cozzini S . A Comprehensive analysis of process energy consumption on multi-socket systems with GPUs ［EB/OL］. ［ 2025-11-17 ］. https：//doi.org/10.48550/arXiv.2409.04941 https://doi.org/10.48550/arXiv.2409.04941 .

Gschwandtner P ， Knobloch M ， Mohr B ， et al . Modeling CPU energy consumption of HPC applications on the IBM POWER ［C］// 2014 22nd Euromicro International Conference on Parallel， Distributed， and Network-Based Processing ， 2014 .

Perron M ， Castro Fernandez R ， Dewitt D ， et al . Cackle： analytical workload cost and performance stability with elastic pools ［J］. Proceedings of the ACM on Management of Data ， 2023 ， 1 （ 4 ）： 1 - 25 .

Kwon G ， Shim Y ， Cho K ， et al . Real-time task scheduling and resource planning for IIoT-based flexible manufacturing with human-machine interaction ［J］. Mathematics ， 2025 ， 13 （ 11 ）： 1842 .

Almaaitah N O ， Singh D E ， Özden T ， et al . Performance-driven scheduling for malleable workloads ［J］. The Journal of Supercomputing ， 2024 ， 80 （ 8 ）： 11556 - 11584 .

Ma J H ， Yao R Y ， Zhang B C ， et al . Data-driven flexibility capability modeling of Internet data center considering task dependency ［J］. IEEE Internet of Things Journal ， 2024 ， 11 （ 14 ）： 24538 - 24550 .

Yalamanchili M T ， Hiremath P ， Mulpuri K . A comprehensive survey on AI-enhanced CPU scheduling in real-time environments： techniques， challenges， and opportunities ［J］. International Research Journal of Engineering and Technology ， 2024 ， 11 （ 12 ）： 785 - 794 .

Gu Y ， Liu Z ， Dai S ， et al . Deep reinforcement learning for job scheduling and resource management in cloud computing： an algorithm-level review ［EB/OL］. ［ 2025-11-17 ］. https：//doi.org/10.48550/arXiv.2501.01007 https://doi.org/10.48550/arXiv.2501.01007 .

开放数据中心标准推进委员会 . 数据中心储能白皮书［R］. 北京：开放数据中心标准推进委员会， 2019 .

Open Data Center Committee （ODCC） . Data center energy storage white paper ［R］. Beijing ： Open Data Center Committee ， 2019 .

Zhang Y ， Tsiligkaridis A ， Paschalidis I C ， et al . Data center and load aggregator coordination towards electricity demand response ［J］. Sustainable Computing： Informatics and Systems ， 2024 ， 42 ： 100957 .

Zhang X ， Yang T ， Yan M ， et al . Demand response participation of data centers with QoS guarantees： hierarchical scheduling and server power capping ［C］// IEEE International Conference on High Performance Computing ， 2020 .

龚向阳，李晓光，张红，等 . 宽带通信网原理［M］. 北京：北京邮电大学出版社， 2006 .

Gong Xiangyang ， Li Xiaoguang ， Zhang Hong ， et al . Principles of broadband communication networks ［M］. Beijing ： Beijing University of Posts and Telecommunications Press ， 2006 .

Hennessy J L ， Patterson D A . Computer architecture： a quantitative approach ［M］. San Francisco， USA ： Morgan Kaufmann Publishers Inc. ， 2022 .

Bakhvalov D . Performance analysis and tuning on modern CPUs ［EB/OL］. ［ 2025-11-17 ］. https：//faculty.cs.niu.edu/~winans/notes/patmc.pdf https://faculty.cs.niu.edu/~winans/notes/patmc.pdf .

荒井英辅，邵春林，蔡凤鸣 . 集成电路［M］. 北京：科学出版社， 2000 .

Arai Eisuke ， Shao Chunlin ， Cai Fengming . Integrated circuits ［M］. Beijing ： Science Press ， 2000 .

Lawrence Livermore National Laboratory . Introduction to parallel computing tutorial ［EB/OL］. ［ 2025-11-17 ］. https：//hpc.llnl.gov/documentation/tutorials/introduction-parallel-computing-tutorial https://hpc.llnl.gov/documentation/tutorials/introduction-parallel-computing-tutorial .

Adefemi T . What every computer scientist needs to know about parallelization ［EB/OL］. ［ 2025-11-17 ］. https：//arxiv.org/abs/2504.03647 https://arxiv.org/abs/2504.03647 .

TU Delft OCW . 4.2.4 Job structure in distributed systems workloads ［EB/OL］. ［ 2025-11-17 ］. https：//ocw.tudelft.nl/course-readings/4-2-4-job-structure-in-distributed-systems-work loads/ https://ocw.tudelft.nl/course-readings/4-2-4-job-structure-in-distributed-systems-workloads/ .

Luo Aijia ， Chen Min ， Wang Yi ， et al . Load flexibility modeling of CPU-intensive computing tasks in data centers ［EB/OL］. ［ 2025-11-17 ］. https：//doi.org/10.36227/techrxiv.176003095.52850256/v1 https://doi.org/10.36227/techrxiv.176003095.52850256/v1 .

Liu Y C ， Qu H X ， Chen S ， et al . Energy efficient task scheduling for heterogeneous multicore processors in edge computing ［J］. Scientific Reports ， 2025 ， 15 ： 11819 .

Hahnel M ， Arega F M ， Dargie W ， et al . Application interference analysis： towards energy-efficient workload management on heterogeneous micro-server architectures ［C］// 2017 IEEE Conference on Computer Communications Workshops ， 2017 .

Standard Performance Evaluation Corporation （SPEC） . SPEC CPU 2017 benchmark suite ［EB/OL］. ［ 2025-11-17 ］. https：//www.spec.org/cpu2017 https://www.spec.org/cpu2017 .

Striani S ， Unterluggauer T ， Andersen P B ， et al . Flexibility potential quantification of electric vehicle charging clusters ［J］. Sustainable Energy， Grids and Networks ， 2024 ， 40 ： 101547 .

国网浙江省电力有限公司 . 1 . 5 倍代理购电价格表［EB/OL］.［ 2025-11-17 ］. https：//www.sx.gov.cn/art/2024/11/28/art_1229366601_4199797.html https://www.sx.gov.cn/art/2024/11/28/art_1229366601_4199797.html .

State Grid Zhejiang Electric Power Co.， Ltd . 1.5x agency electricity purchase price list ［EB/OL］. ［ 2025-11-17 ］. https：//www.sx.gov.cn/art/2024/11/28/art_1229366601_4199797.html https://www.sx.gov.cn/art/2024/11/28/art_1229366601_4199797.html .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Game Trading Strategy for Data Center Aggregators in Medium- and Long-term and Spot Markets

Key Technical Challenges and Prospects of Active Grid Support from Uninterruptible Power Supply of Data Centers

Computing Power-Electric Power Coordinated Dispatch Method for Wide-area Data Center Clusters Based on Spatial Flexibility Characterization

Green Power Directly-connected Data Centers: Coupling Logic, Key Technologies, and Application Modes

Data-driven Distributionally Robust Optimization and Operation Method for Active Distribution Networks Based on Kolmogorov-Arnold Networks

Related Author

ZHENG Fangfang

SUN Qiuye

HUANG Yujia

LI Yushuai

WANG Yingwei

GUAN Yue

SONG Meng

SUN Yaofei

Related Institution

School of Electrical Engineering, Shenyang University of Technology

School of Electrical Engineering, Northeastern University

School of Computer Science, Aalborg University, Aalborg

School of Electrical Engineering, Southeast University

State Grid Liaoning Electric Power Co., Ltd.

AI问答

⁰