{"id":"https://openalex.org/W7155509943","doi":"https://doi.org/10.1145/3767295.3803603","title":"HARP: Orchestrating Automated Parallel Training on Heterogeneous GPU Clusters","display_name":"HARP: Orchestrating Automated Parallel Training on Heterogeneous GPU Clusters","publication_year":2026,"publication_date":"2026-04-24","ids":{"openalex":"https://openalex.org/W7155509943","doi":"https://doi.org/10.1145/3767295.3803603"},"language":null,"primary_location":{"id":"doi:10.1145/3767295.3803603","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803603","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3767295.3803603","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5005459311","display_name":"Antian Liang","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Antian Liang","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0003-2841-7334","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078367414","display_name":"Zhigang Zhao","orcid":"https://orcid.org/0000-0002-4144-8587"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhigang Zhao","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-4144-8587","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065782115","display_name":"K Zhang","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kai Zhang","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0001-7518-5466","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086770018","display_name":"XJ Shi","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xuri Shi","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0009-4451-8550","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069802703","display_name":"Chuantao Li","orcid":"https://orcid.org/0000-0002-4917-7750"},"institutions":[{"id":"https://openalex.org/I4210158984","display_name":"National Supercomputing Center in Wuxi","ror":"https://ror.org/04ypjrs34","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210158984"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chuantao Li","raw_affiliation_strings":["Shandong Computer Science Center (National Supercomputer Center in Jinan), Jinan, China"],"raw_orcid":"https://orcid.org/0009-0000-2725-7277","affiliations":[{"raw_affiliation_string":"Shandong Computer Science Center (National Supercomputer Center in Jinan), Jinan, China","institution_ids":["https://openalex.org/I4210158984"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5057152743","display_name":"Chunxiao Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I4210158984","display_name":"National Supercomputing Center in Wuxi","ror":"https://ror.org/04ypjrs34","country_code":"CN","type":"other","lineage":["https://openalex.org/I4210158984"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chunxiao Wang","raw_affiliation_strings":["Shandong Computer Science Center (National Supercomputer Center in Jinan), Jinan, China"],"raw_orcid":"https://orcid.org/0009-0000-6617-0924","affiliations":[{"raw_affiliation_string":"Shandong Computer Science Center (National Supercomputer Center in Jinan), Jinan, China","institution_ids":["https://openalex.org/I4210158984"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134541658","display_name":"Zhenying He","orcid":"https://orcid.org/0000-0002-2926-4814"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenying He","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-2926-4814","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087411191","display_name":"Yinan Jing","orcid":"https://orcid.org/0000-0002-1169-8032"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yinan Jing","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-1169-8032","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5134557843","display_name":"X. Sean Wang","orcid":"https://orcid.org/0000-0002-9059-3713"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"X. Sean Wang","raw_affiliation_strings":["Fudan University, Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-9059-3713","affiliations":[{"raw_affiliation_string":"Fudan University, Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.47467685,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1160","last_page":"1180"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12292","display_name":"Graph Theory and Algorithms","score":0.18850000202655792,"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"}},"topics":[{"id":"https://openalex.org/T12292","display_name":"Graph Theory and Algorithms","score":0.18850000202655792,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.16179999709129333,"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/T12535","display_name":"Machine Learning and Data Classification","score":0.04450000077486038,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.5270000100135803},{"id":"https://openalex.org/keywords/feature","display_name":"Feature (linguistics)","score":0.2671999931335449},{"id":"https://openalex.org/keywords/training-set","display_name":"Training set","score":0.2574000060558319},{"id":"https://openalex.org/keywords/identification","display_name":"Identification (biology)","score":0.2540999948978424}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7200000286102295},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.5270000100135803},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3653999865055084},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3154999911785126},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.3073999881744385},{"id":"https://openalex.org/C2776401178","wikidata":"https://www.wikidata.org/wiki/Q12050496","display_name":"Feature (linguistics)","level":2,"score":0.2671999931335449},{"id":"https://openalex.org/C459310","wikidata":"https://www.wikidata.org/wiki/Q117801","display_name":"Computational science","level":1,"score":0.25940001010894775},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2574000060558319},{"id":"https://openalex.org/C116834253","wikidata":"https://www.wikidata.org/wiki/Q2039217","display_name":"Identification (biology)","level":2,"score":0.2540999948978424},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.24570000171661377}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3767295.3803603","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803603","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2509.24859","is_oa":true,"landing_page_url":"https://arxiv.org/abs/2509.24859","pdf_url":"https://arxiv.org/pdf/2509.24859","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3767295.3803603","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3767295.3803603","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 21st European Conference on Computer Systems","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W2969388332","https://openalex.org/W3129831491","https://openalex.org/W3204998121","https://openalex.org/W3206832494","https://openalex.org/W4385597806","https://openalex.org/W4394946008","https://openalex.org/W4409363181","https://openalex.org/W4414736061","https://openalex.org/W7133212984","https://openalex.org/W7133227460"],"related_works":[],"abstract_inverted_index":{"With":[0],"the":[1,7,81,109],"rapid":[2],"evolution":[3],"of":[4,9,114],"GPU":[5],"architectures,":[6],"heterogeneity":[8],"model":[10,29],"training":[11,62,148],"infrastructures":[12],"is":[13],"steadily":[14],"increasing.":[15],"In":[16,53,98],"such":[17],"environments,":[18],"effectively":[19],"utilizing":[20],"all":[21],"available":[22],"heterogeneous":[23,49,67,96,144],"accelerators":[24,50],"becomes":[25],"critical":[26],"for":[27,38,66,80],"distributed":[28],"training.":[30],"However,":[31],"existing":[32],"frameworks,":[33],"which":[34],"are":[35],"primarily":[36],"designed":[37,64],"homogeneous":[39],"clusters,":[40],"often":[41],"exhibit":[42],"significant":[43],"resource":[44],"underutilization":[45],"when":[46],"deployed":[47],"on":[48,117,143],"and":[51,112],"networks.":[52],"this":[54],"paper,":[55],"we":[56],"present":[57],"Harp,":[58],"an":[59],"automated":[60],"parallel":[61,83],"framework":[63],"specifically":[65],"clusters.":[68],"Harp":[69,86,100,137],"introduces":[70],"a":[71,77,102],"fine-grained":[72],"planner":[73],"that":[74,106,136],"efficiently":[75],"searches":[76],"wide":[78],"space":[79],"inter-operator":[82],"strategy,":[84],"enabling":[85],"to":[87],"alleviate":[88],"communication":[89],"overheads":[90],"while":[91,126],"maintaining":[92],"balanced":[93],"loads":[94],"across":[95],"accelerators.":[97],"addition,":[99],"implements":[101],"heterogeneity-aware":[103],"1F1B":[104],"scheduler":[105],"adaptively":[107],"adjusts":[108],"execution":[110],"timing":[111],"ordering":[113],"microbatches":[115],"based":[116],"network":[118],"characteristics,":[119],"maximizing":[120],"computation-communication":[121],"overlap":[122],"under":[123],"cross-cluster":[124],"interconnects":[125],"incurring":[127],"only":[128],"minimal":[129],"memory":[130],"overhead.":[131],"Our":[132],"evaluation":[133],"results":[134],"show":[135],"can":[138],"deliver":[139],"1.3\u00d7\u20131.6\u00d7":[140],"higher":[141],"performance":[142],"clusters":[145],"than":[146],"state-of-the-art":[147],"frameworks.":[149]},"counts_by_year":[],"updated_date":"2026-08-05T07:39:15.569665","created_date":"2026-04-25T00:00:00"}