{"id":"https://openalex.org/W4414210491","doi":"https://doi.org/10.1145/3731545.3731581","title":"SAFusion: Efficient Tensor Fusion with Sparsification Ahead for High-Performance Distributed DNN Training","display_name":"SAFusion: Efficient Tensor Fusion with Sparsification Ahead for High-Performance Distributed DNN Training","publication_year":2025,"publication_date":"2025-07-20","ids":{"openalex":"https://openalex.org/W4414210491","doi":"https://doi.org/10.1145/3731545.3731581"},"language":"en","primary_location":{"id":"doi:10.1145/3731545.3731581","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3731545.3731581","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 34th 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/A5029299398","display_name":"Zhangqiang Ming","orcid":"https://orcid.org/0000-0003-1616-8054"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Zhangqiang Ming","raw_affiliation_strings":["Huazhong University of Science and Technology, Wuhan, Hubei, China"],"raw_orcid":"https://orcid.org/0000-0003-1616-8054","affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology, Wuhan, Hubei, China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065831190","display_name":"Yuchong Hu","orcid":"https://orcid.org/0000-0003-1265-7141"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yuchong Hu","raw_affiliation_strings":["Huazhong University of Science and Technology, Wuhan, Hubei, China","Shenzhen Research Institute of Huazhong University of Science and Technology, Shenzhen, Guangdong, China"],"raw_orcid":"https://orcid.org/0000-0003-1265-7141","affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology, Wuhan, Hubei, China","institution_ids":["https://openalex.org/I47720641"]},{"raw_affiliation_string":"Shenzhen Research Institute of Huazhong University of Science and Technology, Shenzhen, Guangdong, China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100882939","display_name":"Xin Zheng","orcid":"https://orcid.org/0009-0007-7321-9264"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xinjue Zheng","raw_affiliation_strings":["Huazhong University of Science and Technology, Wuhan, Hubei, China"],"raw_orcid":"https://orcid.org/0009-0007-7321-9264","affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology, Wuhan, Hubei, China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112600223","display_name":"Wenxiang Zhou","orcid":"https://orcid.org/0009-0007-6391-7798"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Wenxiang Zhou","raw_affiliation_strings":["Huazhong University of Science and Technology, Wuhan, Hubei, China"],"raw_orcid":"https://orcid.org/0009-0007-6391-7798","affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology, Wuhan, Hubei, China","institution_ids":["https://openalex.org/I47720641"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5057421680","display_name":"Dan Feng","orcid":"https://orcid.org/0000-0002-4674-6006"},"institutions":[{"id":"https://openalex.org/I47720641","display_name":"Huazhong University of Science and Technology","ror":"https://ror.org/00p991c53","country_code":"CN","type":"education","lineage":["https://openalex.org/I47720641"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dan Feng","raw_affiliation_strings":["Huazhong University of Science and Technology, Wuhan, Hubei, China"],"raw_orcid":"https://orcid.org/0000-0002-4674-6006","affiliations":[{"raw_affiliation_string":"Huazhong University of Science and Technology, Wuhan, Hubei, China","institution_ids":["https://openalex.org/I47720641"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":5,"corresponding_author_ids":["https://openalex.org/A5029299398"],"corresponding_institution_ids":["https://openalex.org/I47720641"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.22716561,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"14"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9987999796867371,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9987999796867371,"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/T11775","display_name":"COVID-19 diagnosis using AI","score":0.9866999983787537,"subfield":{"id":"https://openalex.org/subfields/2741","display_name":"Radiology, Nuclear Medicine and Imaging"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}},{"id":"https://openalex.org/T12702","display_name":"Brain Tumor Detection and Classification","score":0.9764000177383423,"subfield":{"id":"https://openalex.org/subfields/2808","display_name":"Neurology"},"field":{"id":"https://openalex.org/fields/28","display_name":"Neuroscience"},"domain":{"id":"https://openalex.org/domains/1","display_name":"Life Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.6399999856948853},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.6150000095367432},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.58160001039505},{"id":"https://openalex.org/keywords/synchronizing","display_name":"Synchronizing","score":0.4959000051021576},{"id":"https://openalex.org/keywords/training","display_name":"Training (meteorology)","score":0.46000000834465027},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4542999863624573},{"id":"https://openalex.org/keywords/gradient-method","display_name":"Gradient method","score":0.43540000915527344}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6729000210762024},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.6399999856948853},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.6150000095367432},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.58160001039505},{"id":"https://openalex.org/C162932704","wikidata":"https://www.wikidata.org/wiki/Q1058791","display_name":"Synchronizing","level":3,"score":0.4959000051021576},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46389999985694885},{"id":"https://openalex.org/C2777211547","wikidata":"https://www.wikidata.org/wiki/Q17141490","display_name":"Training (meteorology)","level":2,"score":0.46000000834465027},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4542999863624573},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.44690001010894775},{"id":"https://openalex.org/C115680565","wikidata":"https://www.wikidata.org/wiki/Q5977448","display_name":"Gradient method","level":2,"score":0.43540000915527344},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4092999994754791},{"id":"https://openalex.org/C158525013","wikidata":"https://www.wikidata.org/wiki/Q2593739","display_name":"Fusion","level":2,"score":0.39590001106262207},{"id":"https://openalex.org/C33954974","wikidata":"https://www.wikidata.org/wiki/Q486494","display_name":"Sensor fusion","level":2,"score":0.3714999854564667},{"id":"https://openalex.org/C132459708","wikidata":"https://www.wikidata.org/wiki/Q744069","display_name":"Extrapolation","level":2,"score":0.3619000017642975},{"id":"https://openalex.org/C113315163","wikidata":"https://www.wikidata.org/wiki/Q7625159","display_name":"Structure tensor","level":3,"score":0.28369998931884766},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.2773999869823456}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3731545.3731581","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3731545.3731581","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 34th International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":24,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W2108598243","https://openalex.org/W2194775991","https://openalex.org/W2580688187","https://openalex.org/W2745269232","https://openalex.org/W2886288284","https://openalex.org/W2963786636","https://openalex.org/W2966527647","https://openalex.org/W2975712713","https://openalex.org/W2982475424","https://openalex.org/W2985108934","https://openalex.org/W3047357290","https://openalex.org/W3081168214","https://openalex.org/W3086105743","https://openalex.org/W3091097978","https://openalex.org/W3177554119","https://openalex.org/W3193985311","https://openalex.org/W3204921648","https://openalex.org/W3206636350","https://openalex.org/W4214711850","https://openalex.org/W4221008227","https://openalex.org/W4224307896","https://openalex.org/W4312768957","https://openalex.org/W4402048013"],"related_works":[],"abstract_inverted_index":{"Distributed":[0],"deep":[1],"neural":[2],"networks":[3],"(DNN)":[4],"training":[5],"systems":[6],"deployed":[7],"across":[8],"workers":[9,22],"have":[10],"been":[11],"widely":[12],"used":[13],"in":[14],"various":[15],"domains,":[16],"while":[17],"the":[18,29,49,57,79,116,120],"communication":[19,34,58,80],"overhead":[20],"among":[21],"for":[23],"synchronizing":[24],"gradient":[25,44,50,67,91,111],"tensors":[26,68,112],"often":[27,38,89],"becomes":[28],"performance":[30],"bottleneck.":[31],"To":[32],"optimize":[33],"efficiency,":[35],"state-of-the-art":[36],"studies":[37,88],"apply":[39,90],"both":[40],"two":[41],"techniques:":[42],"i)":[43],"sparsification":[45,92],"compression,":[46],"which":[47,64,101],"truncates":[48],"to":[51,55,73,77,103],"its":[52],"largest":[53],"elements":[54],"reduce":[56,78],"traffic,":[59],"and":[60],"ii)":[61],"tensor":[62,94,99],"fusion,":[63],"merges":[65],"multiple":[66],"within":[69],"a":[70,104,107],"fusion":[71,95],"buffer":[72],"transmit":[74],"them":[75],"together":[76],"startup":[81],"overhead.":[82],"However,":[83],"we":[84],"find":[85],"that":[86,106],"existing":[87],"after":[93,115],"(we":[96],"call":[97],"sparsification-behind":[98],"fusion),":[100],"leads":[102],"fact":[105],"lot":[108],"of":[109],"fused":[110],"are":[113],"missed":[114],"sparsification,":[117],"thus":[118],"impairing":[119],"convergence":[121],"performance.":[122]},"counts_by_year":[],"updated_date":"2025-12-28T23:10:05.387466","created_date":"2025-10-10T00:00:00"}