{"id":"https://openalex.org/W4385623050","doi":"https://doi.org/10.1145/3588195.3592990","title":"Redundancy-Free High-Performance Dynamic GNN Training with Hierarchical Pipeline Parallelism","display_name":"Redundancy-Free High-Performance Dynamic GNN Training with Hierarchical Pipeline Parallelism","publication_year":2023,"publication_date":"2023-08-07","ids":{"openalex":"https://openalex.org/W4385623050","doi":"https://doi.org/10.1145/3588195.3592990"},"language":"en","primary_location":{"id":"doi:10.1145/3588195.3592990","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3588195.3592990","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-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":"https://openalex.org/A5038795287","display_name":"Yaqi Xia","orcid":"https://orcid.org/0009-0006-8101-785X"},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Yaqi Xia","raw_affiliation_strings":["Wuhan University, Wuhan, China"],"affiliations":[{"raw_affiliation_string":"Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013177309","display_name":"Zheng Zhang","orcid":"https://orcid.org/0000-0001-6599-9976"},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zheng Zhang","raw_affiliation_strings":["Wuhan University, Wuhan, China"],"affiliations":[{"raw_affiliation_string":"Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5036278941","display_name":"H. Wang","orcid":"https://orcid.org/0009-0003-8329-0145"},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hulin Wang","raw_affiliation_strings":["Wuhan University, Wuhan, China"],"affiliations":[{"raw_affiliation_string":"Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005205340","display_name":"Donglin Yang","orcid":"https://orcid.org/0000-0002-3913-3623"},"institutions":[{"id":"https://openalex.org/I4210127875","display_name":"Nvidia (United States)","ror":"https://ror.org/03jdj4y14","country_code":"US","type":"company","lineage":["https://openalex.org/I4210127875"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Donglin Yang","raw_affiliation_strings":["Nvidia Corporation, Santa Clara, CA, USA"],"affiliations":[{"raw_affiliation_string":"Nvidia Corporation, Santa Clara, CA, USA","institution_ids":["https://openalex.org/I4210127875"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101677080","display_name":"Xiaobo Zhou","orcid":"https://orcid.org/0009-0004-9500-3390"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Xiaobo Zhou","raw_affiliation_strings":["University of Macau, Macau, China"],"affiliations":[{"raw_affiliation_string":"University of Macau, Macau, China","institution_ids":["https://openalex.org/I204512498"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5063911669","display_name":"Dazhao Cheng","orcid":"https://orcid.org/0000-0003-2869-7623"},"institutions":[{"id":"https://openalex.org/I37461747","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17","country_code":"CN","type":"education","lineage":["https://openalex.org/I37461747"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dazhao Cheng","raw_affiliation_strings":["Wuhan University, Wuhan, China"],"affiliations":[{"raw_affiliation_string":"Wuhan University, Wuhan, China","institution_ids":["https://openalex.org/I37461747"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5038795287"],"corresponding_institution_ids":["https://openalex.org/I37461747"],"apc_list":null,"apc_paid":null,"fwci":2.9574,"has_fulltext":false,"cited_by_count":17,"citation_normalized_percentile":{"value":0.92803848,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"17","last_page":"30"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11273","display_name":"Advanced Graph Neural Networks","score":0.9995999932289124,"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/T11273","display_name":"Advanced Graph Neural Networks","score":0.9995999932289124,"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/T12292","display_name":"Graph Theory and Algorithms","score":0.9941999912261963,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9876999855041504,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8842811584472656},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5842540264129639},{"id":"https://openalex.org/keywords/redundancy","display_name":"Redundancy (engineering)","score":0.5417280197143555},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.5370624661445618},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5108256340026855},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.49865198135375977},{"id":"https://openalex.org/keywords/load-balancing","display_name":"Load balancing (electrical power)","score":0.41755300760269165},{"id":"https://openalex.org/keywords/performance-improvement","display_name":"Performance improvement","score":0.4125659763813019},{"id":"https://openalex.org/keywords/computer-network","display_name":"Computer network","score":0.1942116618156433},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.11098802089691162}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8842811584472656},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5842540264129639},{"id":"https://openalex.org/C152124472","wikidata":"https://www.wikidata.org/wiki/Q1204361","display_name":"Redundancy (engineering)","level":2,"score":0.5417280197143555},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5370624661445618},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5108256340026855},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.49865198135375977},{"id":"https://openalex.org/C138959212","wikidata":"https://www.wikidata.org/wiki/Q1806783","display_name":"Load balancing (electrical power)","level":3,"score":0.41755300760269165},{"id":"https://openalex.org/C2778915421","wikidata":"https://www.wikidata.org/wiki/Q3643177","display_name":"Performance improvement","level":2,"score":0.4125659763813019},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.1942116618156433},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.11098802089691162},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.0},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C21547014","wikidata":"https://www.wikidata.org/wiki/Q1423657","display_name":"Operations management","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3588195.3592990","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3588195.3592990","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 32nd International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320324116","display_name":"Wuhan University","ror":"https://ror.org/033vjfk17"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":31,"referenced_works":["https://openalex.org/W1994616650","https://openalex.org/W2018346931","https://openalex.org/W2064675550","https://openalex.org/W2558460151","https://openalex.org/W2605185701","https://openalex.org/W2734601503","https://openalex.org/W2888096830","https://openalex.org/W2965683718","https://openalex.org/W2970929262","https://openalex.org/W3037699692","https://openalex.org/W3086105743","https://openalex.org/W3086238199","https://openalex.org/W3109841242","https://openalex.org/W3125626040","https://openalex.org/W3131369229","https://openalex.org/W3132185085","https://openalex.org/W3133028082","https://openalex.org/W3157805807","https://openalex.org/W3159109662","https://openalex.org/W3159953606","https://openalex.org/W3166605255","https://openalex.org/W3200735485","https://openalex.org/W3206504463","https://openalex.org/W3210361503","https://openalex.org/W4281398238","https://openalex.org/W4281725510","https://openalex.org/W4283366315","https://openalex.org/W4288070868","https://openalex.org/W4290944486","https://openalex.org/W4321446145","https://openalex.org/W6839594614"],"related_works":["https://openalex.org/W2058965144","https://openalex.org/W2164382479","https://openalex.org/W2146343568","https://openalex.org/W98480971","https://openalex.org/W2150291671","https://openalex.org/W4246549241","https://openalex.org/W2168758875","https://openalex.org/W2410733619","https://openalex.org/W2963483475","https://openalex.org/W4384937382"],"abstract_inverted_index":{"Temporal":[0],"Graph":[1,8],"Neural":[2,9],"Networks(TGNNs)":[3],"extend":[4],"the":[5,41,71,102,112,132,136,141,164],"success":[6],"of":[7,82,87,151,170],"Networks":[10],"to":[11,25,39,130,160,172],"dynamic":[12,88,109],"graphs.":[13],"Distributed":[14],"TGNN":[15,67,83],"training":[16,68],"requires":[17],"efficiently":[18],"tackling":[19],"temporal":[20],"dependency,":[21],"which":[22],"often":[23],"leads":[24],"excessive":[26],"cross-device":[27],"communication":[28,52,105,133,174],"that":[29,100,155],"generates":[30],"significant":[31],"redundant":[32,103],"data.":[33],"However,":[34],"existing":[35],"systems":[36],"are":[37],"unable":[38],"remove":[40],"redundancy":[42],"in":[43,54],"data":[44,94,104,123],"reuse":[45],"and":[46,48,64,85,96,106,127,167],"transfer,":[47],"suffer":[49],"from":[50],"severe":[51],"overhead":[53],"a":[55,118,168],"distributed":[56],"setting.":[57],"This":[58],"paper":[59],"presents":[60],"Sven,":[61],"an":[62,148],"algorithm":[63],"system":[65],"co-designed":[66],"library":[69],"for":[70],"end-to-end":[72],"performance":[73],"optimization":[74],"on":[75,140,147],"multi-node":[76],"multi-GPU":[77],"systems.":[78],"Exploiting":[79],"dependency":[80],"patterns":[81],"models":[84],"characteristics":[86],"graph":[89],"datasets,":[90],"we":[91,116],"design":[92],"redundancy-free":[93],"organization":[95],"load-balancing":[97],"partitioning":[98],"strategies":[99],"mitigate":[101,131],"evenly":[107],"partition":[108],"graphs":[110],"at":[111],"vertex":[113],"level.":[114],"Furthermore,":[115],"develop":[117],"hierarchical":[119],"pipeline":[120],"mechanism":[121],"integrating":[122],"prefetching,":[124],"micro-batch":[125],"pipelining,":[126],"asynchronous":[128],"pipelining":[129],"overhead.":[134],"As":[135],"first":[137],"scaling":[138],"study":[139],"memory-based":[142],"TGNNs":[143],"training,":[144],"experiments":[145],"conducted":[146],"HPC":[149],"cluster":[150],"64":[152],"GPUs":[153],"show":[154],"Sven":[156],"can":[157],"achieve":[158],"up":[159,171],"1.7x-3.3x":[161],"speedup":[162],"over":[163],"state-of-art":[165],"approaches":[166],"factor":[169],"5.26x":[173],"efficiency":[175],"improvement.":[176]},"counts_by_year":[{"year":2025,"cited_by_count":9},{"year":2024,"cited_by_count":6},{"year":2023,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}