{"id":"https://openalex.org/W7125984064","doi":"https://doi.org/10.1109/ton.2026.3658194","title":"Rail: ReArranging Inter-GPU Links for GPU-Centric Clusters","display_name":"Rail: ReArranging Inter-GPU Links for GPU-Centric Clusters","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7125984064","doi":"https://doi.org/10.1109/ton.2026.3658194"},"language":null,"primary_location":{"id":"doi:10.1109/ton.2026.3658194","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ton.2026.3658194","pdf_url":null,"source":{"id":"https://openalex.org/S5407042750","display_name":"IEEE Transactions on Networking","issn_l":"2998-4157","issn":["2998-4157"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Networking","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":"https://openalex.org/A5116800071","display_name":"Haixin Nan","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Haixin Nan","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0009-0007-8741-0625","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078133753","display_name":"Jun Xu","orcid":"https://orcid.org/0000-0003-1494-8129"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jun Xu","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065763589","display_name":"P. Cao","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Peirui Cao","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-0222-4943","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101183397","display_name":"Zhaochen Zhang","orcid":"https://orcid.org/0000-0002-5725-3661"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhaochen Zhang","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-5725-3661","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100633632","display_name":"Yue Wang","orcid":"https://orcid.org/0000-0003-0098-5359"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yizhi Wang","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124139671","display_name":"Zhehao Lin","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhehao Lin","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5108908888","display_name":"Yuhang Li","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuhang Li","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103558100","display_name":"Chong Huang","orcid":"https://orcid.org/0000-0002-6079-6579"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chengyuan Huang","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026477293","display_name":"Wentao Fan","orcid":"https://orcid.org/0000-0001-6694-7289"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wentao Fan","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124095719","display_name":"Xiaohu Xu","orcid":null},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiaohu Xu","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090945517","display_name":"Zhongming Ji","orcid":"https://orcid.org/0000-0001-9732-6338"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhongming Ji","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":"https://orcid.org/0000-0001-9732-6338","affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041184548","display_name":"Shengju Zhang","orcid":"https://orcid.org/0009-0009-0486-1709"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shengju Zhang","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100433001","display_name":"Dongxu Wang","orcid":"https://orcid.org/0000-0002-4427-4412"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dongxu Wang","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101322682","display_name":"Lingkun Meng","orcid":"https://orcid.org/0009-0007-9905-6328"},"institutions":[{"id":"https://openalex.org/I180662265","display_name":"China Mobile (China)","ror":"https://ror.org/05gftfe97","country_code":"CN","type":"company","lineage":["https://openalex.org/I180662265"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Lingkun Meng","raw_affiliation_strings":["China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China"],"raw_orcid":"https://orcid.org/0009-0007-9905-6328","affiliations":[{"raw_affiliation_string":"China Mobile (Suzhou) Software Technology Company Ltd., Suzhou, China","institution_ids":["https://openalex.org/I180662265"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123976250","display_name":"Rong Gu","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Rong Gu","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-1565-9997","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123241526","display_name":"Guihai Chen","orcid":null},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Guihai Chen","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0002-6934-1685","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"last","author":{"id":null,"display_name":"Chen Tian","orcid":"https://orcid.org/0000-0003-2710-7628"},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chen Tian","raw_affiliation_strings":["State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China"],"raw_orcid":"https://orcid.org/0000-0003-2710-7628","affiliations":[{"raw_affiliation_string":"State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China","institution_ids":["https://openalex.org/I881766915"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":17,"corresponding_author_ids":["https://openalex.org/A5116800071"],"corresponding_institution_ids":["https://openalex.org/I881766915"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.13406429,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"34","issue":null,"first_page":"3061","last_page":"3074"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.4350000023841858,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.4350000023841858,"subfield":{"id":"https://openalex.org/subfields/1708","display_name":"Hardware and Architecture"},"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":0.11900000274181366,"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/T12292","display_name":"Graph Theory and Algorithms","score":0.09830000251531601,"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/bottleneck","display_name":"Bottleneck","score":0.8263000249862671},{"id":"https://openalex.org/keywords/bandwidth","display_name":"Bandwidth (computing)","score":0.6162999868392944},{"id":"https://openalex.org/keywords/network-congestion","display_name":"Network congestion","score":0.49059998989105225},{"id":"https://openalex.org/keywords/partition","display_name":"Partition (number theory)","score":0.427700012922287},{"id":"https://openalex.org/keywords/inefficiency","display_name":"Inefficiency","score":0.41780000925064087},{"id":"https://openalex.org/keywords/telecommunications-network","display_name":"Telecommunications network","score":0.41780000925064087},{"id":"https://openalex.org/keywords/domain","display_name":"Domain (mathematical analysis)","score":0.4072999954223633},{"id":"https://openalex.org/keywords/transmission","display_name":"Transmission (telecommunications)","score":0.36230000853538513}],"concepts":[{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.8263000249862671},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.7692000269889832},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.73580002784729},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.6248000264167786},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.6162999868392944},{"id":"https://openalex.org/C195563490","wikidata":"https://www.wikidata.org/wiki/Q180368","display_name":"Network congestion","level":3,"score":0.49059998989105225},{"id":"https://openalex.org/C42812","wikidata":"https://www.wikidata.org/wiki/Q1082910","display_name":"Partition (number theory)","level":2,"score":0.427700012922287},{"id":"https://openalex.org/C192126672","wikidata":"https://www.wikidata.org/wiki/Q1068715","display_name":"Telecommunications network","level":2,"score":0.41780000925064087},{"id":"https://openalex.org/C2778869765","wikidata":"https://www.wikidata.org/wiki/Q6028363","display_name":"Inefficiency","level":2,"score":0.41780000925064087},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.4072999954223633},{"id":"https://openalex.org/C761482","wikidata":"https://www.wikidata.org/wiki/Q118093","display_name":"Transmission (telecommunications)","level":2,"score":0.36230000853538513},{"id":"https://openalex.org/C123745756","wikidata":"https://www.wikidata.org/wiki/Q1665949","display_name":"Interconnection","level":2,"score":0.34869998693466187},{"id":"https://openalex.org/C193415008","wikidata":"https://www.wikidata.org/wiki/Q639681","display_name":"Network architecture","level":2,"score":0.33820000290870667},{"id":"https://openalex.org/C88796919","wikidata":"https://www.wikidata.org/wiki/Q1142907","display_name":"Backbone network","level":2,"score":0.2782999873161316},{"id":"https://openalex.org/C5038329","wikidata":"https://www.wikidata.org/wiki/Q1142907","display_name":"Core network","level":2,"score":0.2775999903678894},{"id":"https://openalex.org/C201100257","wikidata":"https://www.wikidata.org/wiki/Q393287","display_name":"Network traffic control","level":3,"score":0.2752000093460083},{"id":"https://openalex.org/C74294265","wikidata":"https://www.wikidata.org/wiki/Q506273","display_name":"Circuit switching","level":2,"score":0.27320000529289246},{"id":"https://openalex.org/C5111613","wikidata":"https://www.wikidata.org/wiki/Q3799443","display_name":"Inter-domain","level":4,"score":0.27059999108314514},{"id":"https://openalex.org/C199845137","wikidata":"https://www.wikidata.org/wiki/Q145490","display_name":"Network topology","level":2,"score":0.26899999380111694},{"id":"https://openalex.org/C203274722","wikidata":"https://www.wikidata.org/wiki/Q7001161","display_name":"Network performance","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C200157131","wikidata":"https://www.wikidata.org/wiki/Q4854763","display_name":"Bandwidth allocation","level":3,"score":0.263700008392334},{"id":"https://openalex.org/C13540734","wikidata":"https://www.wikidata.org/wiki/Q5318996","display_name":"Dynamic network analysis","level":2,"score":0.2606000006198883}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ton.2026.3658194","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ton.2026.3658194","pdf_url":null,"source":{"id":"https://openalex.org/S5407042750","display_name":"IEEE Transactions on Networking","issn_l":"2998-4157","issn":["2998-4157"],"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Transactions on Networking","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3568657126","display_name":null,"funder_award_id":"BK20243053","funder_id":"https://openalex.org/F4320322769","funder_display_name":"Natural Science Foundation of Jiangsu Province"},{"id":"https://openalex.org/G5507134251","display_name":null,"funder_award_id":"U25B2035","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G6965235922","display_name":null,"funder_award_id":"62502194","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"},{"id":"https://openalex.org/G7905192568","display_name":null,"funder_award_id":"62325205","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320322769","display_name":"Natural Science Foundation of Jiangsu Province","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":25,"referenced_works":["https://openalex.org/W2057332538","https://openalex.org/W2111366547","https://openalex.org/W2163404313","https://openalex.org/W2748014115","https://openalex.org/W2926767350","https://openalex.org/W2968108410","https://openalex.org/W2981114289","https://openalex.org/W3016212306","https://openalex.org/W3044837714","https://openalex.org/W3046470751","https://openalex.org/W3072623287","https://openalex.org/W3095319910","https://openalex.org/W3117684406","https://openalex.org/W3129831491","https://openalex.org/W3157657667","https://openalex.org/W4281790033","https://openalex.org/W4391341581","https://openalex.org/W4394923649","https://openalex.org/W4395073435","https://openalex.org/W4401176799","https://openalex.org/W4401176820","https://openalex.org/W4401176822","https://openalex.org/W4413756477","https://openalex.org/W4413757568","https://openalex.org/W4413978010"],"related_works":[],"abstract_inverted_index":{"In":[0],"modern":[1],"GPU-centric":[2],"clusters,":[3],"large-scale":[4],"AI":[5],"training":[6],"relies":[7],"on":[8,158],"two":[9],"distinct":[10,130],"communication":[11,117,139,190,215],"domains:":[12],"a":[13,23,39,59,143,160,209],"high-bandwidth":[14,55],"intra-node":[15,56,89,111,172],"domain":[16,27,57,173,195],"using":[17],"proprietary":[18,93],"interconnects":[19],"(e.g.,":[20,28],"NVLink),":[21],"and":[22,86,94,207],"scale-out":[24],"inter-node":[25,64,79],"network":[26,65,80,177,194],"RDMA).":[29],"We":[30],"observe":[31],"that":[32,122,148,202],"the":[33,49,53,63,74,110,129,159,170,175,180,193,220],"widely-used":[34,221],"ring":[35],"algorithm,":[36],"often":[37,92],"create":[38],"significant":[40],"load":[41],"imbalance":[42],"across":[43],"these":[44],"domains.":[45],"This":[46,68,119],"leads":[47],"to":[48,97,174,187,219],"counter-intuitive":[50],"scenario":[51],"where":[52],"expensive,":[54],"becomes":[58],"performance":[60],"bottleneck,":[61],"while":[62],"remains":[66],"underutilized.":[67],"inefficiency":[69],"is":[70,82,91],"further":[71],"exacerbated":[72],"by":[73,113],"disparity":[75],"in":[76,213],"bandwidth":[77,81,90,217],"provisioning:":[78],"generally":[83],"more":[84,95],"cost-effective":[85],"accessible,":[87],"whereas":[88],"costly":[96],"scale.":[98],"To":[99],"address":[100],"this":[101],"fundamental":[102],"imbalance,":[103],"we":[104],"propose":[105],"RAIL,":[106],"aimed":[107],"at":[108],"resolving":[109],"bottleneck":[112],"strategically":[114],"rearranging":[115],"inter-GPU":[116],"paths.":[118],"rebalancing":[120],"ensures":[121],"traffic":[123,168],"loads":[124],"are":[125],"appropriately":[126],"matched":[127],"with":[128],"transmission":[131,161],"capabilities":[132,162],"of":[133,163],"each":[134],"domain,":[135],"thereby":[136],"maximizing":[137],"overall":[138],"performance.":[140],"RAIL":[141,203],"incorporates":[142],"Load":[144],"Distributing":[145],"Strategy":[146,182],"(LDS)":[147],"can":[149],"accurately":[150],"partition":[151],"physical":[152],"nodes":[153,156],"into":[154],"logical":[155,197],"based":[157],"both":[164],"domains,":[165],"shifting":[166],"excess":[167],"from":[169],"overloaded":[171],"underutilized":[176],"domain.":[178],"Additionally,":[179],"Intra-Rail":[181],"(IRS)":[183],"leverages":[184],"topological":[185],"characteristics":[186],"ensure":[188],"optimal":[189],"paths":[191],"through":[192],"between":[196],"nodes.":[198],"Our":[199],"evaluation":[200],"demonstrates":[201],"effectively":[204],"mitigates":[205],"congestion":[206],"achieves":[208],"30.7%":[210],"average":[211],"increase":[212],"collective":[214],"bus":[216],"compared":[218],"NCCL":[222],"solution.":[223]},"counts_by_year":[],"updated_date":"2026-02-06T02:01:19.302388","created_date":"2026-01-29T00:00:00"}
