000906547 001__ 906547
000906547 005__ 20250314084120.0
000906547 0247_ $$2doi$$a10.3389/fninf.2021.785068
000906547 0247_ $$2Handle$$a2128/30872
000906547 0247_ $$2altmetric$$aaltmetric:124576629
000906547 0247_ $$2pmid$$a35300490
000906547 0247_ $$2WOS$$aWOS:000773679000001
000906547 037__ $$aFZJ-2022-01510
000906547 082__ $$a610
000906547 1001_ $$0P:(DE-Juel1)165321$$aPronold, Jari$$b0$$eCorresponding author
000906547 245__ $$aRouting Brain Traffic Through the Von Neumann Bottleneck: Parallel Sorting and Refactoring
000906547 260__ $$aLausanne$$bFrontiers Research Foundation$$c2022
000906547 3367_ $$2DRIVER$$aarticle
000906547 3367_ $$2DataCite$$aOutput Types/Journal article
000906547 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1658496564_6740
000906547 3367_ $$2BibTeX$$aARTICLE
000906547 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000906547 3367_ $$00$$2EndNote$$aJournal Article
000906547 520__ $$aGeneric simulation code for spiking neuronal networks spends the major part of the time in the phase where spikes have arrived at a compute node and need to be delivered to their target neurons. These spikes were emitted over the last interval between communication steps by source neurons distributed across many compute nodes and are inherently irregular and unsorted with respect to their targets. For finding those targets, the spikes need to be dispatched to a three-dimensional data structure with decisions on target thread and synapse type to be made on the way. With growing network size, a compute node receives spikes from an increasing number of different source neurons until in the limit each synapse on the compute node has a unique source. Here, we show analytically how this sparsity emerges over the practically relevant range of network sizes from a hundred thousand to a billion neurons. By profiling a production code we investigate opportunities for algorithmic changes to avoid indirections and branching. Every thread hosts an equal share of the neurons on a compute node. In the original algorithm, all threads search through all spikes to pick out the relevant ones. With increasing network size, the fraction of hits remains invariant but the absolute number of rejections grows. Our new alternative algorithm equally divides the spikes among the threads and immediately sorts them in parallel according to target thread and synapse type. After this, every thread completes delivery solely of the section of spikes for its own neurons. Independent of the number of threads, all spikes are looked at only two times. The new algorithm halves the number of instructions in spike delivery which leads to a reduction of simulation time of up to 40 %. Thus, spike delivery is a fully parallelizable process with a single synchronization point and thereby well suited for many-core systems. Our analysis indicates that further progress requires a reduction of the latency that the instructions experience in accessing memory. The study provides the foundation for the exploration of methods of latency hiding like software pipelining and software-induced prefetching.
000906547 536__ $$0G:(DE-HGF)POF4-5234$$a5234 - Emerging NC Architectures (POF4-523)$$cPOF4-523$$fPOF IV$$x0
000906547 536__ $$0G:(EU-Grant)785907$$aHBP SGA2 - Human Brain Project Specific Grant Agreement 2 (785907)$$c785907$$fH2020-SGA-FETFLAG-HBP-2017$$x1
000906547 536__ $$0G:(EU-Grant)945539$$aHBP SGA3 - Human Brain Project Specific Grant Agreement 3 (945539)$$c945539$$fH2020-SGA-FETFLAG-HBP-2019$$x2
000906547 536__ $$0G:(EU-Grant)754304$$aDEEP-EST - DEEP - Extreme Scale Technologies (754304)$$c754304$$fH2020-FETHPC-2016$$x3
000906547 536__ $$0G:(DE-HGF)SO-092$$aACA - Advanced Computing Architectures (SO-092)$$cSO-092$$x4
000906547 536__ $$0G:(GEPRIS)368482240$$aGRK 2416:  MultiSenses-MultiScales: Novel approaches to decipher neural processing in multisensory integration (368482240)$$c368482240$$x5
000906547 536__ $$0G:(DE-Juel-1)ATMLPP$$aATMLPP - ATML Parallel Performance (ATMLPP)$$cATMLPP$$x6
000906547 588__ $$aDataset connected to CrossRef, Journals: juser.fz-juelich.de
000906547 7001_ $$0P:(DE-HGF)0$$aJordan, Jakob$$b1
000906547 7001_ $$0P:(DE-Juel1)132302$$aWylie, Brian J. N.$$b2$$ufzj
000906547 7001_ $$0P:(DE-Juel1)187457$$aKitayama, Itaru$$b3$$ufzj
000906547 7001_ $$0P:(DE-Juel1)144174$$aDiesmann, Markus$$b4$$ufzj
000906547 7001_ $$0P:(DE-Juel1)187422$$aKunkel, Susanne$$b5$$ufzj
000906547 773__ $$0PERI:(DE-600)2452979-5$$a10.3389/fninf.2021.785068$$gVol. 15, p. 785068$$p785068$$tFrontiers in neuroinformatics$$v15$$x1662-5196$$y2022
000906547 8564_ $$uhttps://juser.fz-juelich.de/record/906547/files/pronold2022.pdf$$yOpenAccess
000906547 8767_ $$d2022-12-20$$eAPC$$jDeposit$$lDeposit: Frontiers$$z2507,50 USD
000906547 909CO $$ooai:juser.fz-juelich.de:906547$$pdnbdelivery$$popenCost$$pec_fundedresources$$pVDB$$pdriver$$pOpenAPC$$popen_access$$popenaire
000906547 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)165321$$aForschungszentrum Jülich$$b0$$kFZJ
000906547 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)132302$$aForschungszentrum Jülich$$b2$$kFZJ
000906547 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)187457$$aForschungszentrum Jülich$$b3$$kFZJ
000906547 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144174$$aForschungszentrum Jülich$$b4$$kFZJ
000906547 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)187422$$aForschungszentrum Jülich$$b5$$kFZJ
000906547 9131_ $$0G:(DE-HGF)POF4-523$$1G:(DE-HGF)POF4-520$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5234$$aDE-HGF$$bKey Technologies$$lNatural, Artificial and Cognitive Information Processing$$vNeuromorphic Computing and Network Dynamics$$x0
000906547 9141_ $$y2022
000906547 915pc $$0PC:(DE-HGF)0001$$2APC$$aLocal Funding
000906547 915pc $$0PC:(DE-HGF)0002$$2APC$$aDFG OA Publikationskosten
000906547 915pc $$0PC:(DE-HGF)0003$$2APC$$aDOAJ Journal
000906547 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2021-05-04
000906547 915__ $$0StatID:(DE-HGF)1190$$2StatID$$aDBCoverage$$bBiological Abstracts$$d2021-05-04
000906547 915__ $$0LIC:(DE-HGF)CCBY4$$2HGFVOC$$aCreative Commons Attribution CC BY 4.0
000906547 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000906547 915__ $$0StatID:(DE-HGF)0700$$2StatID$$aFees$$d2021-05-04
000906547 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2021-05-04
000906547 915__ $$0StatID:(DE-HGF)0561$$2StatID$$aArticle Processing Charges$$d2021-05-04
000906547 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bFRONT NEUROINFORM : 2021$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)0501$$2StatID$$aDBCoverage$$bDOAJ Seal$$d2021-05-11T13:08:14Z
000906547 915__ $$0StatID:(DE-HGF)0500$$2StatID$$aDBCoverage$$bDOAJ$$d2021-05-11T13:08:14Z
000906547 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bDOAJ : Blind peer review$$d2021-05-11T13:08:14Z
000906547 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)1050$$2StatID$$aDBCoverage$$bBIOSIS Previews$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)1110$$2StatID$$aDBCoverage$$bCurrent Contents - Clinical Medicine$$d2022-11-24
000906547 915__ $$0StatID:(DE-HGF)9900$$2StatID$$aIF < 5$$d2022-11-24
000906547 9201_ $$0I:(DE-Juel1)INM-6-20090406$$kINM-6$$lComputational and Systems Neuroscience$$x0
000906547 9201_ $$0I:(DE-Juel1)IAS-6-20130828$$kIAS-6$$lTheoretical Neuroscience$$x1
000906547 9201_ $$0I:(DE-Juel1)INM-10-20170113$$kINM-10$$lJara-Institut Brain structure-function relationships$$x2
000906547 9801_ $$aFullTexts
000906547 980__ $$ajournal
000906547 980__ $$aVDB
000906547 980__ $$aI:(DE-Juel1)INM-6-20090406
000906547 980__ $$aI:(DE-Juel1)IAS-6-20130828
000906547 980__ $$aI:(DE-Juel1)INM-10-20170113
000906547 980__ $$aUNRESTRICTED
000906547 980__ $$aAPC
000906547 981__ $$aI:(DE-Juel1)IAS-6-20130828