{"id":"https://openalex.org/W7082981432","doi":"https://doi.org/10.1145/3749175","title":"NeutronHeter: Optimizing Distributed Graph Neural Network Training for Heterogeneous Clusters","display_name":"NeutronHeter: Optimizing Distributed Graph Neural Network Training for Heterogeneous Clusters","publication_year":2025,"publication_date":"2025-09-22","ids":{"openalex":"https://openalex.org/W7082981432","doi":"https://doi.org/10.1145/3749175"},"language":"en","primary_location":{"id":"doi:10.1145/3749175","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3749175","pdf_url":null,"source":{"id":"https://openalex.org/S4387289859","display_name":"Proceedings of the ACM on Management of Data","issn_l":"2836-6573","issn":["2836-6573"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Management of Data","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Chunyu Cao","orcid":"https://orcid.org/0009-0007-8485-1302"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chunyu Cao","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0007-8485-1302","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xin Ai","orcid":"https://orcid.org/0009-0006-0746-8222"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xin Ai","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0006-0746-8222","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Qiange Wang","orcid":"https://orcid.org/0000-0002-4847-6070"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qiange Wang","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0000-0002-4847-6070","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yanfeng Zhang","orcid":"https://orcid.org/0000-0002-9871-0304"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yanfeng Zhang","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0000-0002-9871-0304","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhenbo Fu","orcid":"https://orcid.org/0009-0009-2317-9561"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenbo Fu","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0009-2317-9561","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Hao Yuan","orcid":"https://orcid.org/0009-0002-6502-7696"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hao Yuan","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0002-6502-7696","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Mingyi Cao","orcid":"https://orcid.org/0009-0007-9646-764X"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mingyi Cao","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0007-9646-764X","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Chaoyi Chen","orcid":"https://orcid.org/0009-0008-5518-9978"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chaoyi Chen","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0009-0008-5518-9978","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yingyou Wen","orcid":"https://orcid.org/0000-0002-6659-1785"},"institutions":[{"id":"https://openalex.org/I4210134419","display_name":"Neusoft (China)","ror":"https://ror.org/02zc84r97","country_code":"CN","type":"company","lineage":["https://openalex.org/I4210134419"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yingyou Wen","raw_affiliation_strings":["Neusoft AI Magic Technology Research, Shenyang, China"],"raw_orcid":"https://orcid.org/0000-0002-6659-1785","affiliations":[{"raw_affiliation_string":"Neusoft AI Magic Technology Research, Shenyang, China","institution_ids":["https://openalex.org/I4210134419"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Yu Gu","orcid":"https://orcid.org/0000-0001-7422-6254"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yu Gu","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0000-0001-7422-6254","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]},{"author_position":"last","author":{"id":null,"display_name":"Ge Yu","orcid":"https://orcid.org/0000-0002-3171-8889"},"institutions":[{"id":"https://openalex.org/I9224756","display_name":"Northeastern University","ror":"https://ror.org/03awzbc87","country_code":"CN","type":"education","lineage":["https://openalex.org/I9224756"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ge Yu","raw_affiliation_strings":["Northeastern University, Shenyang, China"],"raw_orcid":"https://orcid.org/0000-0002-3171-8889","affiliations":[{"raw_affiliation_string":"Northeastern University, Shenyang, China","institution_ids":["https://openalex.org/I9224756"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":2.8481,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.92826977,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":"3","issue":"4","first_page":"1","last_page":"29"},"is_retracted":false,"is_paratext":false,"is_xpac":true,"primary_topic":{"id":"https://openalex.org/T12157","display_name":"Geochemistry and Geologic Mapping","score":0.6927000284194946,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T12157","display_name":"Geochemistry and Geologic Mapping","score":0.6927000284194946,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T13067","display_name":"Geological Modeling and Analysis","score":0.02800000086426735,"subfield":{"id":"https://openalex.org/subfields/1906","display_name":"Geochemistry and Petrology"},"field":{"id":"https://openalex.org/fields/19","display_name":"Earth and Planetary Sciences"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T14311","display_name":"Electrical and Electromagnetic Research","score":0.016300000250339508,"subfield":{"id":"https://openalex.org/subfields/3107","display_name":"Atomic and Molecular Physics, and Optics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/workload","display_name":"Workload","score":0.7124999761581421},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5566999912261963},{"id":"https://openalex.org/keywords/heterogeneous-network","display_name":"Heterogeneous network","score":0.5465999841690063},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.48399999737739563},{"id":"https://openalex.org/keywords/symmetric-multiprocessor-system","display_name":"Symmetric multiprocessor system","score":0.4772999882698059},{"id":"https://openalex.org/keywords/cluster-analysis","display_name":"Cluster analysis","score":0.4738999903202057},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.42750000953674316},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4000000059604645},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.3986000120639801}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8008000254631042},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.7124999761581421},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.6416000127792358},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5566999912261963},{"id":"https://openalex.org/C158207573","wikidata":"https://www.wikidata.org/wiki/Q5747224","display_name":"Heterogeneous network","level":4,"score":0.5465999841690063},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.48399999737739563},{"id":"https://openalex.org/C172430144","wikidata":"https://www.wikidata.org/wiki/Q17111997","display_name":"Symmetric multiprocessor system","level":2,"score":0.4772999882698059},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.4738999903202057},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.42750000953674316},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4000000059604645},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3986000120639801},{"id":"https://openalex.org/C192126672","wikidata":"https://www.wikidata.org/wiki/Q1068715","display_name":"Telecommunications network","level":2,"score":0.3467000126838684},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.3434999883174896},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.33660000562667847},{"id":"https://openalex.org/C179145077","wikidata":"https://www.wikidata.org/wiki/Q5154130","display_name":"Communication complexity","level":2,"score":0.3257000148296356},{"id":"https://openalex.org/C197657726","wikidata":"https://www.wikidata.org/wiki/Q174733","display_name":"Bipartite graph","level":3,"score":0.32440000772476196},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.3172999918460846},{"id":"https://openalex.org/C179799912","wikidata":"https://www.wikidata.org/wiki/Q205084","display_name":"Computational complexity theory","level":2,"score":0.3167000114917755},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.30489999055862427},{"id":"https://openalex.org/C61455927","wikidata":"https://www.wikidata.org/wiki/Q1030529","display_name":"Blossom algorithm","level":3,"score":0.3003999888896942},{"id":"https://openalex.org/C48903430","wikidata":"https://www.wikidata.org/wiki/Q491370","display_name":"Graph partition","level":3,"score":0.29760000109672546},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.2797999978065491},{"id":"https://openalex.org/C101765175","wikidata":"https://www.wikidata.org/wiki/Q577764","display_name":"Communications system","level":2,"score":0.2777000069618225},{"id":"https://openalex.org/C51332947","wikidata":"https://www.wikidata.org/wiki/Q1172305","display_name":"Shared resource","level":2,"score":0.27410000562667847},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.26980000734329224},{"id":"https://openalex.org/C193415008","wikidata":"https://www.wikidata.org/wiki/Q639681","display_name":"Network architecture","level":2,"score":0.26649999618530273},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.263700008392334},{"id":"https://openalex.org/C118993495","wikidata":"https://www.wikidata.org/wiki/Q5042828","display_name":"Electrical efficiency","level":3,"score":0.2615000009536743},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.25679999589920044},{"id":"https://openalex.org/C158156997","wikidata":"https://www.wikidata.org/wiki/Q1416645","display_name":"Models of communication","level":2,"score":0.2549999952316284},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.2547999918460846}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3749175","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3749175","pdf_url":null,"source":{"id":"https://openalex.org/S4387289859","display_name":"Proceedings of the ACM on Management of Data","issn_l":"2836-6573","issn":["2836-6573"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Management of Data","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/10","display_name":"Reduced inequalities","score":0.6313139200210571}],"awards":[{"id":"https://openalex.org/G1547168994","display_name":null,"funder_award_id":"62461146205, U2241212","funder_id":"https://openalex.org/F4320334062","funder_display_name":"National Natural Science Foundation of China-Liaoning Joint Fund"},{"id":"https://openalex.org/G8128513374","display_name":null,"funder_award_id":"2023YFB4503601","funder_id":"https://openalex.org/F4320335777","funder_display_name":"National Key Research and Development Program of China"}],"funders":[{"id":"https://openalex.org/F4320334062","display_name":"National Natural Science Foundation of China-Liaoning Joint Fund","ror":null},{"id":"https://openalex.org/F4320335777","display_name":"National Key Research and Development Program of China","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":41,"referenced_works":["https://openalex.org/W1971630691","https://openalex.org/W1991418309","https://openalex.org/W2066636486","https://openalex.org/W2070232376","https://openalex.org/W2101196063","https://openalex.org/W2118953734","https://openalex.org/W2161455936","https://openalex.org/W2164274587","https://openalex.org/W2166860697","https://openalex.org/W2295258302","https://openalex.org/W2516174802","https://openalex.org/W2612026221","https://openalex.org/W2807021761","https://openalex.org/W2914721378","https://openalex.org/W2970929262","https://openalex.org/W3096566397","https://openalex.org/W3130891725","https://openalex.org/W3133028082","https://openalex.org/W3159109662","https://openalex.org/W4220807331","https://openalex.org/W4281725510","https://openalex.org/W4283314525","https://openalex.org/W4290944486","https://openalex.org/W4312910503","https://openalex.org/W4372267520","https://openalex.org/W4381328668","https://openalex.org/W4381328689","https://openalex.org/W4385270195","https://openalex.org/W4385763851","https://openalex.org/W4386128172","https://openalex.org/W4389115441","https://openalex.org/W4394923015","https://openalex.org/W4394946008","https://openalex.org/W4396571432","https://openalex.org/W4396601317","https://openalex.org/W4399208305","https://openalex.org/W4399403248","https://openalex.org/W4401408735","https://openalex.org/W4402042236","https://openalex.org/W4403334614","https://openalex.org/W4404181172"],"related_works":[],"abstract_inverted_index":{"Distributed":[0],"training":[1,20,96],"is":[2,130],"critical":[3,174],"for":[4,98],"scaling":[5],"graph":[6],"neural":[7],"networks":[8],"(GNNs)":[9],"to":[10,25,33,219,222],"large":[11],"graphs.":[12],"However,":[13],"existing":[14],"systems":[15],"struggle":[16],"with":[17,74,84,155,198],"efficient":[18,94,120],"GNN":[19,31,65,95],"in":[21,47,71,194,207],"heterogeneous":[22,34,72,99,196,208],"clusters":[23,35,73,197,209],"due":[24],"two":[26],"main":[27],"challenges.":[28],"First,":[29,101],"mapping":[30,53,68,107,116,123,143],"workloads":[32,63],"involves":[36],"matching":[37],"computation":[38],"and":[39,61,78,139,180,201],"communication":[40,44,60,76,163,168,183,192],"heterogeneity":[41],"while":[42],"minimizing":[43],"overhead,":[45],"resulting":[46],"a":[48,104,118,125,152,214],"non-trivial":[49],"multi-constrained":[50,113],"multi-way":[51,114],"workload":[52,67,106,115,122,145],"problem.":[54],"Second,":[55,158],"the":[56,111,144],"positive":[57],"correlation":[58],"between":[59],"computational":[62,79,203],"of":[64],"makes":[66],"difficult,":[69],"especially":[70],"asymmetric":[75,195],"bandwidth":[77],"power":[80,138],"(e.g.,":[81],"high-end":[82],"GPUs":[83],"low-bandwidth":[85,199],"networks).":[86],"In":[87],"this":[88],"paper,":[89],"we":[90,102,159],"present":[91],"NeutronHeter,":[92],"an":[93,161],"system":[97],"clusters.":[100],"adopt":[103,160],"multi-level":[105],"framework.":[108],"It":[109],"converts":[110],"original":[112],"into":[117],"more":[119],"top-down":[121],"on":[124,136,177],"tree-like":[126],"resource":[127],"graph,":[128],"which":[129],"constructed":[131],"using":[132],"hierarchical":[133],"clustering":[134],"based":[135],"computing":[137],"bandwidth.":[140],"By":[141],"iteratively":[142],"layer":[146],"by":[147,172],"layer,":[148],"our":[149],"design":[150],"achieves":[151,213],"near-optimal":[153],"solution":[154],"low":[156],"overhead.":[157],"adaptive":[162],"migration":[164],"strategy":[165],"that":[166,211],"reduces":[167,190],"over":[169],"slow":[170],"connections":[171,179],"replicating":[173],"remote":[175],"vertices":[176],"these":[178],"managing":[181],"their":[182],"through":[184],"faster":[185],"connections.":[186],"This":[187],"approach":[188],"significantly":[189],"high":[191,202],"overhead":[193],"links":[200],"power.":[204],"Experimental":[205],"results":[206],"show":[210],"NeutronHeter":[212],"speedup":[215],"ranging":[216],"from":[217],"1.06x":[218],"33.05x":[220],"compared":[221],"SOTA":[223],"frameworks.":[224]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-07-08T08:33:18.762332","created_date":"2025-10-10T00:00:00"}