Contribution to a conference proceedings/Contribution to a book

PUBDB-2026-02254

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Pooling HPC Resources Across Organizations and Reducing Carbon Emissions with Transparent, User-Centric Meta-Scheduling

Horn, R. ; Plaß, M. ; Neumann, P.Extern*DESY*

2026
IEEE
ISBN: 979-8-3195-3432-3

This record in other databases:  

Please use a persistent id in citations: doi:10.1109/ISPDC69862.2026.00012

Abstract: We present a solution to unify various High Performance Computing (HPC) resources and enable environmentallysustainable (“green”), automated job assignment for independentHPC systems. The goal is to increase utilization of carbon-neutralcomputing resources and reduce the time to solution with simplejob management using a job specification file and commandline interface modelled after established batch systems. Ourworkload-agnostic approach is motivated by a comprehensiveoverview of existing software, which reveals the lack of asolution for executing arbitrary jobs for multiple users acrossmultiple targets with low technical requirements and complexity.The proposed solution is transparent to HPC administratorsthrough SSH-based instrumentation and to users with regardto the utilized batch system. We demonstrate the capabilitiesof our prototype in a case study of Molecular Dynamics (MD)simulations. The individual jobs are allocated over four differenttarget systems and the scheduling policy takes into account theavailability of resources which can be powered exclusively byRenewable Energy (RE). This results in a reduction of CarbonEmissions (CEs) by at least 9% compared to running all jobson a conventional HPC system. In the future, we plan to extendour scheduling policy to exploit job execution timing using gridcarbon intensity over time and investigate additional simulationparameters to minimize the economical and ecological impact ofHPC jobs.Index Terms—meta-scheduling, green computing, high performance computing, molecular dynamics

---

Contributing Institute(s):

1. Informationstechnologie (IT)

Research Program(s):

1. 6G9 - IDAF (DESY) (POF4-6G9) (POF4-6G9)

Experiment(s):

1. No specific instrument

Appears in the scientific report 2026

---

---

Linked articles:

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Book/Proceedings Nunez, M. (Editor)Extern* ; Neumann, P. (Editor)Extern*DESY* ; Neuwirth, S. (Editor)Extern* ; Hunold, S. (Editor)Extern* ; Pollinger, T. (Editor)Extern* ; Schreiber, M. (Editor)Extern* ; Servan, S. (Editor)DESY*
International Symposium on Parallel and Distributed Computing: proceedings, ISPDC, 25th conference, Hamburg, Germany, July 1-2, 2026
20262026 25th International Symposium on Parallel and Distributed Computing (ISPDC), ISPDC, HamburgHamburg, Germany, 1 Jul 2026 - 2 Jul 20262026-07-012026-07-02 IEEE (2026)2026     Files BibTeX | EndNote: XML, Text | RIS

---