KIT | KIT-Bibliothek | Impressum | Datenschutz

Experience with ARM WNs at the WLCG Tier1 GridKa

Krull, Armin Alfredo 1; Schnepf, Matthias Jochen 1; Fischer, Max ORCID iD icon 2; Petzold, Andreas 1; Szumlak, T. [Hrsg.]; Rachwał, B. [Hrsg.]; Dziurda, A. [Hrsg.]; Schulz, M. [Hrsg.]
1 Karlsruher Institut für Technologie (KIT)
2 Scientific Computing Center (SCC), Karlsruher Institut für Technologie (KIT)

Abstract:

Computing centers are always looking for new worker nodes (WNs) that can reduce operational costs, especially energy consumption, and provide higher performance. ARM CPUs promise higher power efficiency than x86 CPUs. The WLCG Tier1 center GridKa will therefore partially use WNs with ARM CPUs. It has already conducted various energy consumption and performance benchmarkings for several such ARM WNs based on the benchmark HEPScore23 as well as a CPU frequency scan to determine their best performance and power efficiency. We present the results of these benchmarkings in comparison to those of the x86 WNs currently operated at GridKa and summarize the progress made at GridKa in providing ARM WNs to the HEP community.


Verlagsausgabe §
DOI: 10.5445/IR/1000188062
Veröffentlicht am 05.12.2025
Originalveröffentlichung
DOI: 10.1051/epjconf/202533701147
Cover der Publikation
Zugehörige Institution(en) am KIT Scientific Computing Center (SCC)
Publikationstyp Zeitschriftenaufsatz
Publikationsjahr 2025
Sprache Englisch
Identifikator ISSN: 2100-014X
KITopen-ID: 1000188062
Erschienen in EPJ Web of Conferences
Verlag EDP Sciences
Band 337
Seiten Art.-Nr.: 01147
Bemerkung zur Veröffentlichung 27th International Conference on Computing in High Energy and Nuclear Physics (CHEP 2024), Krakow, 21st-25th October 2024
Vorab online veröffentlicht am 07.10.2025
Nachgewiesen in OpenAlex
Scopus
KIT – Die Universität in der Helmholtz-Gemeinschaft
KITopen Landing Page