Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings

Demetris Trihinas; Lauritz Thamsen

doi:10.1007/978-3-031-48803-0_28

Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings

Demetris Trihinas, Lauritz Thamsen

Department of Computer Science

University of Glasgow

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

As the Internet of Things (IoT) increasingly empowers the network extremes with in-place intelligence through Machine Learning (ML), energy consumption and carbon emissions become crucial factors. ML is often computationally intensive, with state-of-the-art model architectures consuming significant energy per training round and imposing a large carbon footprint. This work, therefore, argues for the need to introduce novel mechanisms into the ML pipelines of IoT services, so that energy awareness is integrated in the decision-making process for when and where to initiate ML model training.

Original language	English
Title of host publication	Euro-Par 2023
Subtitle of host publication	Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers
Editors	Demetris Zeinalipour, Dora Blanco Heras, George Pallis, Herodotos Herodotou, Demetris Trihinas, Daniel Balouek, Patrick Diehl, Terry Cojean, Karl Fürlinger, Maja Hanne Kirkeby, Matteo Nardelli, Pierangelo Di Sanzo
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	256-259
Number of pages	4
ISBN (Print)	9783031488023
DOIs	https://doi.org/10.1007/978-3-031-48803-0_28
Publication status	Published - 2024
Event	International workshops held at the 29th International Conference on Parallel and Distributed Computing, Euro-Par 2023 - Limassol, Cyprus Duration: 28 Aug 2023 → 1 Sept 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14352 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International workshops held at the 29th International Conference on Parallel and Distributed Computing, Euro-Par 2023
Country/Territory	Cyprus
City	Limassol
Period	28/08/23 → 1/09/23

Keywords

Carbon Footprint
Distributed Systems
Energy Profiling
Internet of Things
Machine Learning
System Orchestration

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1007/978-3-031-48803-0_28

Cite this

Trihinas, D., & Thamsen, L. (2024). Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings. In D. Zeinalipour, D. Blanco Heras, G. Pallis, H. Herodotou, D. Trihinas, D. Balouek, P. Diehl, T. Cojean, K. Fürlinger, M. H. Kirkeby, M. Nardelli, & P. Di Sanzo (Eds.), Euro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers (pp. 256-259). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14352 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-48803-0_28

Trihinas, Demetris ; Thamsen, Lauritz. / Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings. Euro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers. editor / Demetris Zeinalipour ; Dora Blanco Heras ; George Pallis ; Herodotos Herodotou ; Demetris Trihinas ; Daniel Balouek ; Patrick Diehl ; Terry Cojean ; Karl Fürlinger ; Maja Hanne Kirkeby ; Matteo Nardelli ; Pierangelo Di Sanzo. Springer Science and Business Media Deutschland GmbH, 2024. pp. 256-259 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0b55d790bdd14d6c9192e35dd24ae095,

title = "Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings",

abstract = "As the Internet of Things (IoT) increasingly empowers the network extremes with in-place intelligence through Machine Learning (ML), energy consumption and carbon emissions become crucial factors. ML is often computationally intensive, with state-of-the-art model architectures consuming significant energy per training round and imposing a large carbon footprint. This work, therefore, argues for the need to introduce novel mechanisms into the ML pipelines of IoT services, so that energy awareness is integrated in the decision-making process for when and where to initiate ML model training.",

keywords = "Carbon Footprint, Distributed Systems, Energy Profiling, Internet of Things, Machine Learning, System Orchestration",

author = "Demetris Trihinas and Lauritz Thamsen",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.; International workshops held at the 29th International Conference on Parallel and Distributed Computing, Euro-Par 2023 ; Conference date: 28-08-2023 Through 01-09-2023",

year = "2024",

doi = "10.1007/978-3-031-48803-0_28",

language = "English",

isbn = "9783031488023",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "256--259",

editor = "Demetris Zeinalipour and {Blanco Heras}, Dora and George Pallis and Herodotos Herodotou and Demetris Trihinas and Daniel Balouek and Patrick Diehl and Terry Cojean and Karl F{\"u}rlinger and Kirkeby, {Maja Hanne} and Matteo Nardelli and {Di Sanzo}, Pierangelo",

booktitle = "Euro-Par 2023",

address = "Germany",

}

Trihinas, D & Thamsen, L 2024, Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings. in D Zeinalipour, D Blanco Heras, G Pallis, H Herodotou, D Trihinas, D Balouek, P Diehl, T Cojean, K Fürlinger, MH Kirkeby, M Nardelli & P Di Sanzo (eds), Euro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14352 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 256-259, International workshops held at the 29th International Conference on Parallel and Distributed Computing, Euro-Par 2023, Limassol, Cyprus, 28/08/23. https://doi.org/10.1007/978-3-031-48803-0_28

Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings. / Trihinas, Demetris; Thamsen, Lauritz.
Euro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers. ed. / Demetris Zeinalipour; Dora Blanco Heras; George Pallis; Herodotos Herodotou; Demetris Trihinas; Daniel Balouek; Patrick Diehl; Terry Cojean; Karl Fürlinger; Maja Hanne Kirkeby; Matteo Nardelli; Pierangelo Di Sanzo. Springer Science and Business Media Deutschland GmbH, 2024. p. 256-259 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14352 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings

AU - Trihinas, Demetris

AU - Thamsen, Lauritz

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.

PY - 2024

Y1 - 2024

N2 - As the Internet of Things (IoT) increasingly empowers the network extremes with in-place intelligence through Machine Learning (ML), energy consumption and carbon emissions become crucial factors. ML is often computationally intensive, with state-of-the-art model architectures consuming significant energy per training round and imposing a large carbon footprint. This work, therefore, argues for the need to introduce novel mechanisms into the ML pipelines of IoT services, so that energy awareness is integrated in the decision-making process for when and where to initiate ML model training.

AB - As the Internet of Things (IoT) increasingly empowers the network extremes with in-place intelligence through Machine Learning (ML), energy consumption and carbon emissions become crucial factors. ML is often computationally intensive, with state-of-the-art model architectures consuming significant energy per training round and imposing a large carbon footprint. This work, therefore, argues for the need to introduce novel mechanisms into the ML pipelines of IoT services, so that energy awareness is integrated in the decision-making process for when and where to initiate ML model training.

KW - Carbon Footprint

KW - Distributed Systems

KW - Energy Profiling

KW - Internet of Things

KW - Machine Learning

KW - System Orchestration

UR - http://www.scopus.com/inward/record.url?scp=85191008316&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-48803-0_28

DO - 10.1007/978-3-031-48803-0_28

M3 - Conference contribution

AN - SCOPUS:85191008316

SN - 9783031488023

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 256

EP - 259

BT - Euro-Par 2023

A2 - Zeinalipour, Demetris

A2 - Blanco Heras, Dora

A2 - Pallis, George

A2 - Herodotou, Herodotos

A2 - Trihinas, Demetris

A2 - Balouek, Daniel

A2 - Diehl, Patrick

A2 - Cojean, Terry

A2 - Fürlinger, Karl

A2 - Kirkeby, Maja Hanne

A2 - Nardelli, Matteo

A2 - Di Sanzo, Pierangelo

PB - Springer Science and Business Media Deutschland GmbH

T2 - International workshops held at the 29th International Conference on Parallel and Distributed Computing, Euro-Par 2023

Y2 - 28 August 2023 through 1 September 2023

ER -

Trihinas D, Thamsen L. Towards Energy-Aware Machine Learning in Geo-Distributed IoT Settings. In Zeinalipour D, Blanco Heras D, Pallis G, Herodotou H, Trihinas D, Balouek D, Diehl P, Cojean T, Fürlinger K, Kirkeby MH, Nardelli M, Di Sanzo P, editors, Euro-Par 2023: Parallel Processing Workshops - Euro-Par 2023 International Workshops, 2023, Revised Selected Papers. Springer Science and Business Media Deutschland GmbH. 2024. p. 256-259. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-48803-0_28