Mutli-Metric based GPU Scoring Method in Kubernetes Environments

Jae Hyung Kim; Woosuk Lee; Sang Hyeop Oh; Hyunsu Jeong; Heonchang Yu; Joon Min Gil

doi:10.1109/SERA65747.2025.11154557

Mutli-Metric based GPU Scoring Method in Kubernetes Environments

Jae Hyung Kim^*
, Woosuk Lee
, Sang Hyeop Oh
, Hyunsu Jeong
, Heonchang Yu
, Joon Min Gil

^*Corresponding author for this work

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

Abstract

This paper proposes a GPU scoring method based on multi-metric of GPUs for efficient utilization of GPU resources in a Kubernetes environment. The proposed method calculates the node score based on various metrics of GPUs in each node. This calculation aims to increase the utilization of CPU resources and optimize the performance of the entire cluster. The experimental results show the possibility of improving the efficiency of pod deployment compared to the existing Kubernetes scheduling method.

Original language	English
Title of host publication	2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings
Editors	Yeong-Tae Song, Mingon Kang, Junghwan Rhee
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	547-549
Number of pages	3
ISBN (Electronic)	9798331565367
DOIs	https://doi.org/10.1109/SERA65747.2025.11154557
Publication status	Published - 2025
Event	23rd IEEE/ACIS International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Las Vegas, United States Duration: 2025 May 29 → 2025 May 31

Publication series

Name	2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings

Conference

Conference	23rd IEEE/ACIS International Conference on Software Engineering Research, Management and Applications, SERA 2025
Country/Territory	United States
City	Las Vegas
Period	25/5/29 → 25/5/31

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

Keywords

GPU scoring
Kubernetes
Multi-metric

ASJC Scopus subject areas

Safety, Risk, Reliability and Quality
Artificial Intelligence
Computer Science Applications
Software
Information Systems and Management

Access to Document

10.1109/SERA65747.2025.11154557

Cite this

Kim, J. H., Lee, W., Oh, S. H., Jeong, H., Yu, H., & Gil, J. M. (2025). Mutli-Metric based GPU Scoring Method in Kubernetes Environments. In Y.-T. Song, M. Kang, & J. Rhee (Eds.), 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings (pp. 547-549). (2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SERA65747.2025.11154557

Mutli-Metric based GPU Scoring Method in Kubernetes Environments. / Kim, Jae Hyung; Lee, Woosuk; Oh, Sang Hyeop et al.
2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings. ed. / Yeong-Tae Song; Mingon Kang; Junghwan Rhee. Institute of Electrical and Electronics Engineers Inc., 2025. p. 547-549 (2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings).

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

Kim, JH, Lee, W, Oh, SH, Jeong, H, Yu, H & Gil, JM 2025, Mutli-Metric based GPU Scoring Method in Kubernetes Environments. in Y-T Song, M Kang & J Rhee (eds), 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings. 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 547-549, 23rd IEEE/ACIS International Conference on Software Engineering Research, Management and Applications, SERA 2025, Las Vegas, United States, 25/5/29. https://doi.org/10.1109/SERA65747.2025.11154557

Kim JH, Lee W, Oh SH, Jeong H, Yu H, Gil JM. Mutli-Metric based GPU Scoring Method in Kubernetes Environments. In Song YT, Kang M, Rhee J, editors, 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2025. p. 547-549. (2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings). doi: 10.1109/SERA65747.2025.11154557

Kim, Jae Hyung ; Lee, Woosuk ; Oh, Sang Hyeop et al. / Mutli-Metric based GPU Scoring Method in Kubernetes Environments. 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings. editor / Yeong-Tae Song ; Mingon Kang ; Junghwan Rhee. Institute of Electrical and Electronics Engineers Inc., 2025. pp. 547-549 (2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings).

@inproceedings{95aae594f85f4bceaf926be4998022ae,

title = "Mutli-Metric based GPU Scoring Method in Kubernetes Environments",

abstract = "This paper proposes a GPU scoring method based on multi-metric of GPUs for efficient utilization of GPU resources in a Kubernetes environment. The proposed method calculates the node score based on various metrics of GPUs in each node. This calculation aims to increase the utilization of CPU resources and optimize the performance of the entire cluster. The experimental results show the possibility of improving the efficiency of pod deployment compared to the existing Kubernetes scheduling method.",

keywords = "GPU scoring, Kubernetes, Multi-metric",

author = "Kim, \{Jae Hyung\} and Woosuk Lee and Oh, \{Sang Hyeop\} and Hyunsu Jeong and Heonchang Yu and Gil, \{Joon Min\}",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 23rd IEEE/ACIS International Conference on Software Engineering Research, Management and Applications, SERA 2025 ; Conference date: 29-05-2025 Through 31-05-2025",

year = "2025",

doi = "10.1109/SERA65747.2025.11154557",

language = "English",

series = "2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "547--549",

editor = "Yeong-Tae Song and Mingon Kang and Junghwan Rhee",

booktitle = "2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings",

}

TY - GEN

T1 - Mutli-Metric based GPU Scoring Method in Kubernetes Environments

AU - Kim, Jae Hyung

AU - Lee, Woosuk

AU - Oh, Sang Hyeop

AU - Jeong, Hyunsu

AU - Yu, Heonchang

AU - Gil, Joon Min

PY - 2025

Y1 - 2025

N2 - This paper proposes a GPU scoring method based on multi-metric of GPUs for efficient utilization of GPU resources in a Kubernetes environment. The proposed method calculates the node score based on various metrics of GPUs in each node. This calculation aims to increase the utilization of CPU resources and optimize the performance of the entire cluster. The experimental results show the possibility of improving the efficiency of pod deployment compared to the existing Kubernetes scheduling method.

AB - This paper proposes a GPU scoring method based on multi-metric of GPUs for efficient utilization of GPU resources in a Kubernetes environment. The proposed method calculates the node score based on various metrics of GPUs in each node. This calculation aims to increase the utilization of CPU resources and optimize the performance of the entire cluster. The experimental results show the possibility of improving the efficiency of pod deployment compared to the existing Kubernetes scheduling method.

KW - GPU scoring

KW - Kubernetes

KW - Multi-metric

UR - https://www.scopus.com/pages/publications/105017739419

U2 - 10.1109/SERA65747.2025.11154557

DO - 10.1109/SERA65747.2025.11154557

M3 - Conference contribution

AN - SCOPUS:105017739419

T3 - 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings

SP - 547

EP - 549

BT - 2025 IEEE/ACIS 23rd International Conference on Software Engineering Research, Management and Applications, SERA 2025 - Proceedings

A2 - Song, Yeong-Tae

A2 - Kang, Mingon

A2 - Rhee, Junghwan

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 23rd IEEE/ACIS International Conference on Software Engineering Research, Management and Applications, SERA 2025

Y2 - 29 May 2025 through 31 May 2025

ER -

Mutli-Metric based GPU Scoring Method in Kubernetes Environments

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this