001033567 001__ 1033567
001033567 005__ 20250822121305.0
001033567 0247_ $$2doi$$a10.1109/SCW63240.2024.00158
001033567 0247_ $$2datacite_doi$$a10.34734/FZJ-2024-06447
001033567 0247_ $$2WOS$$aWOS:001451792300127
001033567 037__ $$aFZJ-2024-06447
001033567 1001_ $$0P:(DE-Juel1)187395$$aJohn, Chelsea Maria$$b0$$ufzj
001033567 1112_ $$aSupercomputing Conference 2024, 2024 International Workshop on Performance, Portability, and Productivity in HPC$$cAtlanta$$d2024-11-17 - 2024-11-22$$gSC24$$wUSA
001033567 245__ $$aPerformance and Power: Systematic Evaluation of AI Workloads on Accelerators with CARAML
001033567 260__ $$c2024
001033567 300__ $$aNan
001033567 3367_ $$2ORCID$$aCONFERENCE_PAPER
001033567 3367_ $$033$$2EndNote$$aConference Paper
001033567 3367_ $$2BibTeX$$aINPROCEEDINGS
001033567 3367_ $$2DRIVER$$aconferenceObject
001033567 3367_ $$2DataCite$$aOutput Types/Conference Paper
001033567 3367_ $$0PUB:(DE-HGF)8$$2PUB:(DE-HGF)$$aContribution to a conference proceedings$$bcontrib$$mcontrib$$s1736168766_29200
001033567 500__ $$aAlso available at: https://arxiv.org/abs/2409.12994
001033567 520__ $$aThe rapid advancement of machine learning (ML) technologies has driven the development of specialized hardware accelerators designed to facilitate more efficient model training. This paper introduces the CARAML benchmark suite, which is employed to assess performance and energy consumption during the training of transformer-based large language models and computer vision models on a range of hardware accelerators, including systems from NVIDIA, AMD, and Graphcore. CARAML provides a compact, automated, extensible, and reproducible framework for assessing the performance and energy of ML workloads across various novel hardware architectures. The design and implementation of CARAML, along with a custom power measurement tool called jpwr, are discussed in detail.
001033567 536__ $$0G:(DE-HGF)POF4-5112$$a5112 - Cross-Domain Algorithms, Tools, Methods Labs (ATMLs) and Research Groups (POF4-511)$$cPOF4-511$$fPOF IV$$x0
001033567 536__ $$0G:(DE-HGF)POF4-5121$$a5121 - Supercomputing & Big Data Facilities (POF4-512)$$cPOF4-512$$fPOF IV$$x1
001033567 536__ $$0G:(DE-Juel-1)68GX21007F$$aOpenGPT-X - Aufbau eines Gaia-X Knotens für große KI-Sprachmodelle und innovative Sprachapplikations-Services; Teilvorhaben: Optimierung und Skalierung auf großen HPC-Systemen (68GX21007F)$$c68GX21007F$$x2
001033567 536__ $$0G:(EU-Grant)955513$$aMAELSTROM - MAchinE Learning for Scalable meTeoROlogy and cliMate (955513)$$c955513$$fH2020-JTI-EuroHPC-2019-1$$x3
001033567 536__ $$0G:(BMBF)16HPC029$$aVerbundprojekt: MAELSTROM - Skalierbarkeit von Anwendungen des Maschinellen Lernens in den Bereichen Wetter und Klimawissenschaften für das zukünftige Supercomputing (16HPC029)$$c16HPC029$$x4
001033567 536__ $$0G:(DE-Juel-1)ATML-X-DEV$$aATML-X-DEV - ATML Accelerating Devices (ATML-X-DEV)$$cATML-X-DEV$$x5
001033567 588__ $$aDataset connected to DataCite
001033567 7001_ $$0P:(DE-Juel1)145478$$aHerten, Andreas$$b1$$eCorresponding author$$ufzj
001033567 7001_ $$0P:(DE-Juel1)192254$$aPenke, Carolin$$b2$$eCorresponding author$$ufzj
001033567 7001_ $$0P:(DE-Juel1)172888$$aNassyr, Stepan$$b3$$eCorresponding author$$ufzj
001033567 773__ $$a10.1109/SCW63240.2024.00158
001033567 8564_ $$uhttps://conferences.computer.org/sc-wpub/
001033567 8564_ $$uhttps://juser.fz-juelich.de/record/1033567/files/CARAMLSlides.pdf$$yOpenAccess
001033567 8564_ $$uhttps://juser.fz-juelich.de/record/1033567/files/CARAML_Bench.pdf$$yOpenAccess
001033567 909CO $$ooai:juser.fz-juelich.de:1033567$$pdnbdelivery$$pec_fundedresources$$pVDB$$pdriver$$popen_access$$popenaire
001033567 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)187395$$aForschungszentrum Jülich$$b0$$kFZJ
001033567 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)145478$$aForschungszentrum Jülich$$b1$$kFZJ
001033567 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)192254$$aForschungszentrum Jülich$$b2$$kFZJ
001033567 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)172888$$aForschungszentrum Jülich$$b3$$kFZJ
001033567 9131_ $$0G:(DE-HGF)POF4-511$$1G:(DE-HGF)POF4-510$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5112$$aDE-HGF$$bKey Technologies$$lEngineering Digital Futures – Supercomputing, Data Management and Information Security for Knowledge and Action$$vEnabling Computational- & Data-Intensive Science and Engineering$$x0
001033567 9131_ $$0G:(DE-HGF)POF4-512$$1G:(DE-HGF)POF4-510$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5121$$aDE-HGF$$bKey Technologies$$lEngineering Digital Futures – Supercomputing, Data Management and Information Security for Knowledge and Action$$vSupercomputing & Big Data Infrastructures$$x1
001033567 9141_ $$y2024
001033567 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
001033567 920__ $$lyes
001033567 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
001033567 980__ $$acontrib
001033567 980__ $$aVDB
001033567 980__ $$aUNRESTRICTED
001033567 980__ $$aI:(DE-Juel1)JSC-20090406
001033567 9801_ $$aFullTexts