{"id":"https://openalex.org/W3015585426","doi":"https://doi.org/10.1109/icassp40776.2020.9053834","title":"Overlap Local-SGD: An Algorithmic Approach to Hide Communication Delays in Distributed SGD","display_name":"Overlap Local-SGD: An Algorithmic Approach to Hide Communication Delays in Distributed SGD","publication_year":2020,"publication_date":"2020-04-09","ids":{"openalex":"https://openalex.org/W3015585426","doi":"https://doi.org/10.1109/icassp40776.2020.9053834","mag":"3015585426"},"language":"en","primary_location":{"id":"doi:10.1109/icassp40776.2020.9053834","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9053834","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","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/A5100398947","display_name":"Jianyu Wang","orcid":"https://orcid.org/0000-0001-6611-8528"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Jianyu Wang","raw_affiliation_strings":["Carnegie Mellon University, Pittsburgh, USA"],"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University, Pittsburgh, USA","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101925135","display_name":"Hao Liang","orcid":"https://orcid.org/0000-0003-4918-3285"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hao Liang","raw_affiliation_strings":["Carnegie Mellon University, Pittsburgh, USA"],"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University, Pittsburgh, USA","institution_ids":["https://openalex.org/I74973139"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5067441201","display_name":"Gauri Joshi","orcid":"https://orcid.org/0000-0002-6372-9697"},"institutions":[{"id":"https://openalex.org/I74973139","display_name":"Carnegie Mellon University","ror":"https://ror.org/05x2bcf33","country_code":"US","type":"education","lineage":["https://openalex.org/I74973139"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gauri Joshi","raw_affiliation_strings":["Carnegie Mellon University, Pittsburgh, USA"],"affiliations":[{"raw_affiliation_string":"Carnegie Mellon University, Pittsburgh, USA","institution_ids":["https://openalex.org/I74973139"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5100398947"],"corresponding_institution_ids":["https://openalex.org/I74973139"],"apc_list":null,"apc_paid":null,"fwci":3.0489,"has_fulltext":false,"cited_by_count":35,"citation_normalized_percentile":{"value":0.92836976,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"8871","last_page":"8875"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9997000098228455,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9997000098228455,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.9984999895095825,"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/T12676","display_name":"Machine Learning and ELM","score":0.9925000071525574,"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/computer-science","display_name":"Computer science","score":0.8289211988449097},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.7569923400878906},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7173451781272888},{"id":"https://openalex.org/keywords/node","display_name":"Node (physics)","score":0.5802851319313049},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.5633683204650879},{"id":"https://openalex.org/keywords/convergence","display_name":"Convergence (economics)","score":0.5471413135528564},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5080604553222656},{"id":"https://openalex.org/keywords/distributed-algorithm","display_name":"Distributed algorithm","score":0.45636385679244995},{"id":"https://openalex.org/keywords/mathematical-proof","display_name":"Mathematical proof","score":0.4490634799003601},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.3673084080219269},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.345894992351532},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.32466059923171997},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.2855892777442932},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.25370028614997864}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8289211988449097},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.7569923400878906},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7173451781272888},{"id":"https://openalex.org/C62611344","wikidata":"https://www.wikidata.org/wiki/Q1062658","display_name":"Node (physics)","level":2,"score":0.5802851319313049},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5633683204650879},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.5471413135528564},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5080604553222656},{"id":"https://openalex.org/C130120984","wikidata":"https://www.wikidata.org/wiki/Q2835898","display_name":"Distributed algorithm","level":2,"score":0.45636385679244995},{"id":"https://openalex.org/C108710211","wikidata":"https://www.wikidata.org/wiki/Q11538","display_name":"Mathematical proof","level":2,"score":0.4490634799003601},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.3673084080219269},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.345894992351532},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.32466059923171997},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2855892777442932},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.25370028614997864},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","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/C66938386","wikidata":"https://www.wikidata.org/wiki/Q633538","display_name":"Structural engineering","level":1,"score":0.0},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"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/C50522688","wikidata":"https://www.wikidata.org/wiki/Q189833","display_name":"Economic growth","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icassp40776.2020.9053834","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icassp40776.2020.9053834","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.6299999952316284,"display_name":"Industry, innovation and infrastructure","id":"https://metadata.un.org/sdg/9"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":45,"referenced_works":["https://openalex.org/W104184427","https://openalex.org/W1522301498","https://openalex.org/W2060393849","https://openalex.org/W2083842231","https://openalex.org/W2194775991","https://openalex.org/W2622263826","https://openalex.org/W2896633576","https://openalex.org/W2900182564","https://openalex.org/W2902410171","https://openalex.org/W2906936451","https://openalex.org/W2946900926","https://openalex.org/W2962771678","https://openalex.org/W2963179579","https://openalex.org/W2963361660","https://openalex.org/W2963807318","https://openalex.org/W2964121744","https://openalex.org/W2964243509","https://openalex.org/W2965373594","https://openalex.org/W2966297898","https://openalex.org/W2970289928","https://openalex.org/W2977517840","https://openalex.org/W2996693845","https://openalex.org/W3037047862","https://openalex.org/W3102816259","https://openalex.org/W3118608800","https://openalex.org/W3212066104","https://openalex.org/W4249253882","https://openalex.org/W4287759323","https://openalex.org/W4292928772","https://openalex.org/W4299802343","https://openalex.org/W4318619660","https://openalex.org/W6604254268","https://openalex.org/W6631190155","https://openalex.org/W6728757088","https://openalex.org/W6739622702","https://openalex.org/W6749257597","https://openalex.org/W6752191696","https://openalex.org/W6754143808","https://openalex.org/W6755251367","https://openalex.org/W6756976665","https://openalex.org/W6757392376","https://openalex.org/W6763138067","https://openalex.org/W6766673545","https://openalex.org/W6768660345","https://openalex.org/W6787972765"],"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/W2013643406","https://openalex.org/W2027972911","https://openalex.org/W2157978810","https://openalex.org/W2955229517","https://openalex.org/W2966297898"],"abstract_inverted_index":{"Distributed":[0],"stochastic":[1],"gradient":[2],"descent":[3],"(SGD)":[4],"is":[5,161],"essential":[6],"for":[7,143],"scaling":[8],"the":[9,21,35,73,83,111,132,144],"machine":[10],"learning":[11],"algorithms":[12],"to":[13,64,71,81],"a":[14,43,124,140],"large":[15],"number":[16],"of":[17,37,122,134,153],"computing":[18],"nodes.":[19],"However,":[20],"infrastructures":[22],"variability":[23],"such":[24],"as":[25,70,86],"high":[26],"communication":[27,66],"delay":[28],"or":[29,46],"random":[30],"node":[31],"slowdown":[32],"greatly":[33],"impedes":[34],"performance":[36],"distributed":[38,74],"SGD":[39],"algorithm,":[40],"especially":[41],"in":[42],"wireless":[44],"system":[45],"sensor":[47],"networks.":[48],"In":[49],"this":[50,90,154],"paper,":[51],"we":[52],"propose":[53],"an":[54,93],"algorithmic":[55],"approach":[56,78],"named":[57],"Overlap":[58,135],"Local-SGD":[59],"(and":[60],"its":[61],"momentum":[62],"variant)":[63],"overlap":[65],"and":[67,159],"computation":[68],"so":[69],"speedup":[72],"training":[75,123],"procedure.":[76],"The":[77],"can":[79],"help":[80],"mitigate":[82],"straggler":[84],"effects":[85],"well.":[87],"We":[88,137],"achieve":[89],"by":[91],"adding":[92],"anchor":[94,113],"model":[95,114],"on":[96,128],"each":[97],"node.":[98],"After":[99],"multiple":[100],"local":[101],"updates,":[102],"locally":[103],"trained":[104],"models":[105],"will":[106],"be":[107],"pulled":[108],"back":[109],"towards":[110],"synchronized":[112],"rather":[115],"than":[116],"communicating":[117],"with":[118,156],"others.":[119],"Experimental":[120],"results":[121],"deep":[125],"neural":[126],"network":[127],"CIFAR-10":[129],"dataset":[130],"demonstrate":[131],"effectiveness":[133],"Local-SGD.":[136],"also":[138],"provide":[139],"convergence":[141],"guarantee":[142],"proposed":[145],"algorithm":[146],"under":[147],"non-convex":[148],"objective":[149],"functions.A":[150],"full":[151],"version":[152],"paper":[155],"additional":[157],"examples":[158],"proofs":[160],"accessible":[162],"at:":[163],"http://andrew.cmu.edu/user/gaurij/overlap_local_SGD.pdf.":[164]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":9},{"year":2023,"cited_by_count":6},{"year":2022,"cited_by_count":6},{"year":2021,"cited_by_count":5},{"year":2020,"cited_by_count":6}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}