000907602 001__ 907602
000907602 005__ 20221109161719.0
000907602 0247_ $$2arXiv$$aarXiv:2205.02491
000907602 0247_ $$2doi$$a10.1145/3539781.3539792
000907602 0247_ $$2Handle$$a2128/31597
000907602 037__ $$aFZJ-2022-02101
000907602 1001_ $$0P:(DE-Juel1)178969$$aWu, Xinzhe$$b0$$ufzj
000907602 1112_ $$aPlatform for Advanced Scientific Computing$$cBasel$$d2022-06-27 - 2022-06-29$$gPASC22$$wSwitzerland
000907602 245__ $$aChASE - A Distributed Hybrid CPU-GPU Eigensolver for Large-scale Hermitian Eigenvalue Problems
000907602 260__ $$bACM New York, NY, USA$$c2022
000907602 29510 $$aProceedings of the Platform for Advanced Scientific Computing Conference - ACM New York, NY, USA, 2022. - ISBN 9781450394109 - doi:10.1145/3539781.3539792
000907602 300__ $$a12 pages
000907602 3367_ $$2ORCID$$aCONFERENCE_PAPER
000907602 3367_ $$033$$2EndNote$$aConference Paper
000907602 3367_ $$2BibTeX$$aINPROCEEDINGS
000907602 3367_ $$2DRIVER$$aconferenceObject
000907602 3367_ $$2DataCite$$aOutput Types/Conference Paper
000907602 3367_ $$0PUB:(DE-HGF)8$$2PUB:(DE-HGF)$$aContribution to a conference proceedings$$bcontrib$$mcontrib$$s1659093693_31246
000907602 3367_ $$0PUB:(DE-HGF)7$$2PUB:(DE-HGF)$$aContribution to a book$$mcontb
000907602 520__ $$aAs modern massively parallel clusters are getting larger with beefier compute nodes, traditional parallel eigensolvers, such as direct solvers, struggle keeping the pace with the hardware evolution and being able to scale efficiently due to additional layers of communication and synchronization. This difficulty is especially important when porting traditional libraries to heterogeneous computing architectures equipped with accelerators, such as Graphics Processing Unit (GPU). Recently, there have been significant scientific contributions to the development of filter-based subspace eigensolver to compute partial eigenspectrum. The simpler structure of these type of algorithms makes for them easier to avoid the communication and synchronization bottlenecks typical of direct solvers. The Chebyshev Accelerated Subspace Eigensolver (ChASE) is a modern subspace eigensolver to compute partial extremal eigenpairs of large-scale Hermitian eigenproblems with the acceleration of a filter based on Chebyshev polynomials. In this work, we extend our previous work on ChASE by adding support for distributed hybrid CPU-multi-GPU computing architectures. Our tests show that ChASE achieves very good scaling performance up to 144 nodes with 526 NVIDIA A100 GPUs in total on dense eigenproblems of size up to $360$k.
000907602 536__ $$0G:(DE-HGF)POF4-5111$$a5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511)$$cPOF4-511$$fPOF IV$$x0
000907602 536__ $$0G:(EU-Grant)823767$$aPRACE-6IP - PRACE 6th Implementation Phase Project (823767)$$c823767$$fH2020-INFRAEDI-2018-1$$x1
000907602 536__ $$0G:(DE-Juel1)SDLQM$$aSimulation and Data Laboratory Quantum Materials (SDLQM) (SDLQM)$$cSDLQM$$fSimulation and Data Laboratory Quantum Materials (SDLQM)$$x2
000907602 588__ $$aDataset connected to DataCite
000907602 7001_ $$0P:(DE-HGF)0$$aDavidovic, Davor$$b1
000907602 7001_ $$0P:(DE-Juel1)169552$$aAchilles, Sebastian$$b2$$ufzj
000907602 7001_ $$0P:(DE-Juel1)144723$$aDi Napoli, Edoardo$$b3$$eCorresponding author$$ufzj
000907602 773__ $$a10.1145/3539781.3539792$$pArticle No.: 9
000907602 8564_ $$uhttps://juser.fz-juelich.de/record/907602/files/2205.02491.pdf$$yOpenAccess
000907602 909CO $$ooai:juser.fz-juelich.de:907602$$pdnbdelivery$$pec_fundedresources$$pVDB$$pdriver$$popen_access$$popenaire
000907602 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)178969$$aForschungszentrum Jülich$$b0$$kFZJ
000907602 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)169552$$aForschungszentrum Jülich$$b2$$kFZJ
000907602 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144723$$aForschungszentrum Jülich$$b3$$kFZJ
000907602 9131_ $$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
000907602 9141_ $$y2022
000907602 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000907602 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
000907602 980__ $$acontrib
000907602 980__ $$aVDB
000907602 980__ $$aUNRESTRICTED
000907602 980__ $$acontb
000907602 980__ $$aI:(DE-Juel1)JSC-20090406
000907602 9801_ $$aFullTexts