Automatic Energy-Efficient Job Scheduling in HPC: A Novel SLURM Plugin Approach

Anders Aaen Springborg, Michele Albano, Samuel Xavier-de-Souza

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

Abstract

This paper presents a novel approach to enable energy-efficient job scheduling in High-Performance Computing (HPC) environments through application-specific energy models. We propose an architecture that decouples scheduling heuristics to a Python plugin of the HPC scheduler SLURM. The approach leverages the principles of Service-Oriented Architecture and Clean Architecture to create a proof-of-concept system that is adaptable for production setups, providing a platform for integrating various energy-efficient scheduling models. We demonstrate the approach in a single-node HPC system with an energy saving of 11% for the High-Performance Conjugate Gradients (HPCG) benchmark, which represents modern applications’ data access patterns and computation. The proposed approach opens up possibilities for more complex setups, such as automatically scheduling jobs when energy is cheap and renewable, a practice already used in companies utilizing HPC.
OriginalsprogEngelsk
TitelProceedings of 2023 SC Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis, SC Workshops 2023
Antal sider8
ForlagAssociation for Computing Machinery
Publikationsdato12 nov. 2023
Sider1831-1838
ISBN (Elektronisk)979-8-4007-0785-8
DOI
StatusUdgivet - 12 nov. 2023
BegivenhedWorkshops of The International Conference on High Performance Computing, Network, Storage, and Analysis 2023 - Denver, USA
Varighed: 12 nov. 202317 nov. 2023
https://sc23.supercomputing.org/program/workshops/

Konference

KonferenceWorkshops of The International Conference on High Performance Computing, Network, Storage, and Analysis 2023
Land/OmrådeUSA
ByDenver
Periode12/11/202317/11/2023
Internetadresse

Fingeraftryk

Dyk ned i forskningsemnerne om 'Automatic Energy-Efficient Job Scheduling in HPC: A Novel SLURM Plugin Approach'. Sammen danner de et unikt fingeraftryk.

Citationsformater