research-article

Open access

Energy and Emissions of Machine Learning on Smartphones vs. the Cloud

Authors:

David Patterson,

Jeffrey M. Gilbert,

Marco Gruteser,

Tenghui ZhuAuthors Info & Claims

Communications of the ACM, Volume 67, Issue 2

Pages 86 - 97

https://doi.org/10.1145/3624719

Published: 25 January 2024 Publication History

All formats PDF

Editorial Notes

The authors found and corrected two arithmetic errors in their article, "Energy and Emissions of Machine Learning on Smartphones vs. the Cloud" after it was printed. The corrections affected numerous derived values in the tables and the discussion but the magnitude of the changes was not sufficient to affect the paper's conclusions. Therefore, ACM has published a Corrected Version of Record (CVoR) on January 25, 2024. For reference purposes, the VoR may still be accessed via the Supplemental Material section on this page. Also included in the supplemental material is a Corrigendum, featuring a detailed explanation from the article's authors on how the errors occurred and the ramifications on their findings.

Abstract

A Google case study finds ML training in the cloud can reduce CO₂e emissions up to 100×.

Supplementary Material

Version of Record for "Energy and Emissions of Machine Learning on Smartphones vs. the Cloud" by Patterson et al., Communications of the ACM, Volume 67, Issue 2 (CACM 67:2). (3624719-vor.pdf)

Download
13.45 MB

Corrigendum to "Energy and Emissions of Machine Learning on Smartphones vs. the Cloud" by Patterson et al., Communications of the ACM, Volume 67, Issue 2 (CACM 67:2). (3624719-corrigendum.pdf)

Download
119.51 KB

References

[1]

Ahmad, R. et al. A survey on energy estimation and power modeling schemes for smartphone applications. Int. J. Commun. Syst. 30, 11 (2017).

[2]

Almeida, M. et al. Smart at what cost? Characterising mobile deep neural networks in the wild. In Proceedings of the 21^st ACM Internet Measurement Conf., 2021, 658--672.

Digital Library

[3]

Alphabet. CDP Climate Change Response, Aug. 2022.

[4]

Apple. Use Clean Energy Charging on your iPhone, June 2023.

[5]

Bankmycell. How Many Smartphones Are In The World? 2022.

[6]

Berger, D. et al. Reducing embodied carbon is important. Computer Architecture Today, 2023.

[7]

BusinessWire. One Billion Smartphones Worldwide Have Wireless Charging. 2021.

[8]

Cai, D. et al. Towards ubiquitous learning: A first measurement of on-device training performance. In Proceedings of the 5^th Intern. Workshop on Embedded and Mobile Deep Learning, 2021, 31--36.

Digital Library

[9]

Carroll, A. and Heiser, G. An analysis of power consumption in a smartphone. USENIX Annual Tech. Conf., 2010, 1--14.

[10]

Department of Energy. Compliance Certification Database. Energy Efficiency and Renewable Energy Appliance and Equipment Standards Program. 2020.

[11]

Department of Energy. Code of Federal Regulations, Title 10, 430, Subpart B, Appendix Y: Uniform Test Method for Measuring the Energy Consumption of Battery Chargers. 2022.

[12]

EPA. Greenhouse Gases Equivalencies Calculator - Calculations and References. 2022.

[13]

Ferreira, D., Dey, A.K. and Kostakos, V. Understanding human-smartphone concerns: A study of battery life. In Proceedings of the 2011 Intern. Conf. Pervasive Computing, 19--33.

[14]

Fletcher, C. Private Communication, May 2023.

[15]

Google Environmental Report 2022; https://bit.ly/3QAVDMs

[16]

Gupta, U. et al. Chasing carbon: The elusive environmental footprint of computing. IEEE Micro 42, 4 (2022), 37--47.

Digital Library

[17]

Gupta, U. et al. ACT: Designing sustainable computer systems with an architectural carbon modeling tool. In Proceedings of the 49^th Annual Intern. Symp. Computer Architecture. June 2022, 784--799.

Digital Library

[18]

Hoque, M.A. et al. Modeling, profiling, and debugging the energy consumption of mobile devices. ACM Computing Surveys 48, 3 (2015), 1--40.

[19]

HPE product carbon footprint---HPE ProLiant DL345 Gen10 Plus server data sheet; https://bit.ly/3QCyM31.

[20]

International Energy Agency. Global Energy & CO2 Status Report 2019, Report Extract Emissions.

[21]

Jang, J.W. et al. Sparsity-aware and re-configurable NPU architecture for Samsung flagship mobile SOC. In Proceedings of the ACM/IEEE 48^th Annual Intern. Symp. Computer Architecture. IEEE, 2021, 15--28.

Digital Library

[22]

Javed, A., Shahid, M.A., Sharif, M. and Yasmin, M. Energy consumption in mobile phones. Intern. J. Computer Network & Info. Security 9, 12 (2017), 18--28.

[23]

Jouppi, N.P. et al. Ten lessons from three generations shaped Google's TPUv4i. In Proceedings of the ACM/IEEE 48^th Annual Intern. Symp. Computer Architecture. IEEE, 2021, 1--14.

Digital Library

[24]

Koningstein, R. We Now Do More Computing Where There's Cleaner Energy, May 2021.

[25]

Lövehagen, N., Malmodin, J., Bergmark, P. and Matinfar, S. Assessing embodied carbon emissions of communication user devices by combining approaches. Renewable and Sustainable Energy Reviews 183 (2023), 113422.

[26]

Moen, O.M. et al. Screening life cycle assessment of a new datacenter in Trondheim. SINTEF Report, 2022.

[27]

Meier, A. Personal Communication, June 25, 2022.

[28]

Patterson, D. et al. The carbon footprint of machine learning training will plateau, then shrink. IEEE Computer 55, 7 (2022).

Digital Library

[29]

Perrucci, G.P., Fitzek, F.H. and Widmer, J. Survey on energy consumption entities on the smartphone platform. In Proceedings of the 2011 IEEE 73^rd Vehicular Technology Conf., 1--6.

[30]

Popa, R. Private Communication, May 2023.

[31]

Pramanik, P.K.D. et al. Power consumption analysis, measurement, management, and issues: A state-of-the-art review of smartphone battery and energy usage. IEEE Access 7 (2019), 182113--182172.

[32]

Qiu, X. et al. A First Look Into the Carbon Footprint of Federated Learning. 2021; arXiv:2102.07627

[33]

Saraev, A., Gama, M., Piontek, F. and Negi, P. LCA of Dell Servers R6515, R7515, R6525, R7525. 2021.

[34]

Schwartz, R., Dodge, J., Smith, N.A. and Etzioni, O. Green AI. Commun. ACM 63, 12 (Dec. 2020), 54--63.

Digital Library

[35]

Sun, L., Sheshadri, R., Zheng, W. and Koutsonikolas, D. Modeling WiFi active power/energy consumption in smartphones. In Proceedings of the IEEE 34^th Intern. Conf. Distributed Computing Systems, 2014, 41--51

Digital Library

[36]

Taiwan Semiconductor Manufacturing Company. TSMC 2022 Sustainability Report. 2022

[37]

Thompson, N.C., Greenewald, K., Lee, K. and Manso, G.F. Deep learning's diminishing returns: The cost of improvement is becoming unsustainable. IEEE Spectrum 58, 10 (2021), 50--55.

[38]

Walker, A. We Asked, You Told Us: Most Readers Own Plenty of Smartphone Chargers. 2020.

Cited By

Vahdat AMa XPatterson D(2024)New Computer Evaluation Metrics for a Changing WorldCommunications of the ACM10.1145/3637867Online publication date: 3-Jul-2024
https://dl.acm.org/doi/10.1145/3637867
Kshetri N(2024)The Environmental Impact of Artificial IntelligenceIT Professional10.1109/MITP.2024.339947126:3(9-13)Online publication date: May-2024
https://doi.org/10.1109/MITP.2024.3399471

Index Terms

Energy and Emissions of Machine Learning on Smartphones vs. the Cloud

Recommendations

Prediction on Net Carbon Emissions of Yunnan under Emissions and Forest Carbon Sinks
CDCIEM '11: Proceedings of the 2011 International Conference on Computer Distributed Control and Intelligent Environmental Monitoring

This paper predicts Net Carbon Emissions of Yunnan form 2007 to 2050 through energy consumption carbon emissions based on economic growth model and forest carbon sinks based on CO2FIX model. Further, we analyze the influencing factors of carbon ...
Read More
Measuring regional efficiency of energy and carbon dioxide emissions in China

There exist multiple randomness errors (commonly regarded as the uncertainty) in the estimation of CO2 emissions. This uncertainty has been an important issue in regional energy use and carbon dioxide (CO2) emissions efficiency evaluation. To address ...
Read More
Photovoltaic Energy Prediction Using Machine Learning Techniques
Advances in Computational Intelligence
Abstract
Solar energy is becoming one of the most promising power sources in residential, commercial, and industrial applications. Solar photovoltaic (PV) facilities use PV cells that convert solar irradiation into electric power. PV cells can be used in ...
Read More

Comments

Information & Contributors

Information

Published In

cover image Communications of the ACM

Communications of the ACM Volume 67, Issue 2

February 2024

110 pages

ISSN:0001-0782

EISSN:1557-7317

DOI:10.1145/3641526

Editor:
James Larus
Association for Computing Machinery, New York, NY

Issue’s Table of Contents

Copyright © 2024 Owner/Author.

Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the Owner/Author.

Publisher

Association for Computing Machinery

New York, NY, United States

Publication History

Published: 25 January 2024

Published in CACM Volume 67, Issue 2

Check for updates

Qualifiers

Research-article

Contributors

Other Metrics

View Article Metrics

Bibliometrics & Citations

Bibliometrics

Article Metrics

2
Total Citations
View Citations
4,338
Total Downloads

Downloads (Last 12 months)4,338
Downloads (Last 6 weeks)383

Other Metrics

View Author Metrics

Citations

Cited By

Vahdat AMa XPatterson D(2024)New Computer Evaluation Metrics for a Changing WorldCommunications of the ACM10.1145/3637867Online publication date: 3-Jul-2024
https://dl.acm.org/doi/10.1145/3637867
Kshetri N(2024)The Environmental Impact of Artificial IntelligenceIT Professional10.1109/MITP.2024.339947126:3(9-13)Online publication date: May-2024
https://doi.org/10.1109/MITP.2024.3399471

View Options

View options

PDF

View or Download as a PDF file.

eReader

View online with eReader.

Digital Edition

View this article in digital edition.

Digital Edition

Magazine Site

View this article on the magazine site (external)

Get Access

Login options

Check if you have access through your login credentials or your institution to get full access on this article.

Full Access

Get this Article

Media

Figures

Other

Tables

View Issue’s Table of Contents