001019351 001__ 1019351
001019351 005__ 20231219202011.0
001019351 0247_ $$2doi$$a10.1145/3624062.3624249
001019351 0247_ $$2datacite_doi$$a10.34734/FZJ-2023-05321
001019351 037__ $$aFZJ-2023-05321
001019351 1001_ $$0P:(DE-Juel1)178969$$aWu, Xinzhe$$b0$$eCorresponding author$$ufzj
001019351 1112_ $$aSC-W 2023: Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis$$cDenver, CO$$d2023-11-12 - 2023-11-17$$wUSA
001019351 245__ $$aAdvancing the distributed Multi-GPU ChASE library through algorithm optimization and NCCL library
001019351 260__ $$bACM New York, NY, USA$$c2023
001019351 300__ $$a1688–1696
001019351 3367_ $$2ORCID$$aCONFERENCE_PAPER
001019351 3367_ $$033$$2EndNote$$aConference Paper
001019351 3367_ $$2BibTeX$$aINPROCEEDINGS
001019351 3367_ $$2DRIVER$$aconferenceObject
001019351 3367_ $$2DataCite$$aOutput Types/Conference Paper
001019351 3367_ $$0PUB:(DE-HGF)8$$2PUB:(DE-HGF)$$aContribution to a conference proceedings$$bcontrib$$mcontrib$$s1702991423_20585
001019351 520__ $$aAs supercomputers become larger with powerful Graphics Processing Unit (GPU), traditional direct eigensolvers struggle to keep up with the hardware evolution and scale efficiently due to communication and synchronization demands. Conversely, subspace eigensolvers, like the Chebyshev Accelerated Subspace Eigensolver (ChASE), have a simpler structure and can overcome communication and synchronization bottlenecks. ChASE is a modern subspace eigensolver that uses Chebyshev polynomials to accelerate the computation of extremal eigenpairs of dense Hermitian eigenproblems. In this work we show how we have modified ChASE by rethinking its memory layout, introducing a novel parallelization scheme, switching to a more performing communication-avoiding algorithm for one of its inner modules, and substituting the MPI library by the vendor-optimized NCCL library. The resulting library can tackle dense problems with size up to , and scales effortlessly up to the full 900 nodes—each one powered by 4 × A100 NVIDIA GPUs—of the JUWELS Booster hosted at the Jülich Supercomputing Centre.
001019351 536__ $$0G:(DE-HGF)POF4-5111$$a5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511)$$cPOF4-511$$fPOF IV$$x0
001019351 536__ $$0G:(DE-Juel1)SDLQM$$aSimulation and Data Laboratory Quantum Materials (SDLQM) (SDLQM)$$cSDLQM$$fSimulation and Data Laboratory Quantum Materials (SDLQM)$$x1
001019351 588__ $$aDataset connected to CrossRef Conference
001019351 7001_ $$0P:(DE-Juel1)144723$$aDi Napoli, Edoardo$$b1
001019351 773__ $$a10.1145/3624062.3624249
001019351 8564_ $$uhttps://dl.acm.org/doi/10.1145/3624062.3624249
001019351 8564_ $$uhttps://juser.fz-juelich.de/record/1019351/files/ArXiv%20preprint.pdf$$yOpenAccess
001019351 8564_ $$uhttps://juser.fz-juelich.de/record/1019351/files/ArXiv%20preprint.gif?subformat=icon$$xicon$$yOpenAccess
001019351 8564_ $$uhttps://juser.fz-juelich.de/record/1019351/files/ArXiv%20preprint.jpg?subformat=icon-1440$$xicon-1440$$yOpenAccess
001019351 8564_ $$uhttps://juser.fz-juelich.de/record/1019351/files/ArXiv%20preprint.jpg?subformat=icon-180$$xicon-180$$yOpenAccess
001019351 8564_ $$uhttps://juser.fz-juelich.de/record/1019351/files/ArXiv%20preprint.jpg?subformat=icon-640$$xicon-640$$yOpenAccess
001019351 909CO $$ooai:juser.fz-juelich.de:1019351$$popenaire$$popen_access$$pVDB$$pdriver$$pdnbdelivery
001019351 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)178969$$aForschungszentrum Jülich$$b0$$kFZJ
001019351 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144723$$aForschungszentrum Jülich$$b1$$kFZJ
001019351 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)POF4-5111$$aDE-HGF$$bKey Technologies$$lEngineering Digital Futures – Supercomputing, Data Management and Information Security for Knowledge and Action$$vEnabling Computational- & Data-Intensive Science and Engineering$$x0
001019351 9141_ $$y2023
001019351 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
001019351 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
001019351 9201_ $$0I:(DE-Juel1)CASA-20230315$$kCASA$$lCenter for Advanced Simulation and Analytics$$x1
001019351 980__ $$acontrib
001019351 980__ $$aVDB
001019351 980__ $$aUNRESTRICTED
001019351 980__ $$aI:(DE-Juel1)JSC-20090406
001019351 980__ $$aI:(DE-Juel1)CASA-20230315
001019351 9801_ $$aFullTexts