{"id":"https://openalex.org/W4372267133","doi":"https://doi.org/10.1145/3552326.3567505","title":"Hi-Speed DNN Training with Espresso: Unleashing the Full Potential of Gradient Compression with Near-Optimal Usage Strategies","display_name":"Hi-Speed DNN Training with Espresso: Unleashing the Full Potential of Gradient Compression with Near-Optimal Usage Strategies","publication_year":2023,"publication_date":"2023-05-05","ids":{"openalex":"https://openalex.org/W4372267133","doi":"https://doi.org/10.1145/3552326.3567505"},"language":"en","primary_location":{"id":"doi:10.1145/3552326.3567505","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3552326.3567505","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Eighteenth European Conference on Computer Systems","raw_type":"proceedings-article"},"type":"conference-paper","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/A5100697510","display_name":"Zhuang Wang","orcid":"https://orcid.org/0000-0002-1614-0601"},"institutions":[{"id":"https://openalex.org/I74775410","display_name":"Rice University","ror":"https://ror.org/008zs3103","country_code":"US","type":"education","lineage":["https://openalex.org/I74775410"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhuang Wang","raw_affiliation_strings":["Computer Science Department, Rice University, Houston, Texas, United States of America"],"raw_orcid":"https://orcid.org/0000-0002-1614-0601","affiliations":[{"raw_affiliation_string":"Computer Science Department, Rice University, Houston, Texas, United States of America","institution_ids":["https://openalex.org/I74775410"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084047386","display_name":"Haibin Lin","orcid":"https://orcid.org/0000-0003-4879-5335"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Haibin Lin","raw_affiliation_strings":["ByteDance Inc., Newark, California, USA"],"raw_orcid":"https://orcid.org/0000-0003-4879-5335","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Newark, California, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028120333","display_name":"Yibo Zhu","orcid":"https://orcid.org/0000-0002-9113-2660"},"institutions":[{"id":"https://openalex.org/I4210108985","display_name":"Bellevue Hospital Center","ror":"https://ror.org/01ky34z31","country_code":"US","type":"healthcare","lineage":["https://openalex.org/I1283621791","https://openalex.org/I4210086933","https://openalex.org/I4210108985"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yibo Zhu","raw_affiliation_strings":["ByteDance Inc., Bellevue, Washington, United States of America"],"raw_orcid":"https://orcid.org/0000-0002-9113-2660","affiliations":[{"raw_affiliation_string":"ByteDance Inc., Bellevue, Washington, United States of America","institution_ids":["https://openalex.org/I4210108985"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5075120727","display_name":"T. S. Eugene Ng","orcid":"https://orcid.org/0000-0003-2954-0767"},"institutions":[{"id":"https://openalex.org/I74775410","display_name":"Rice University","ror":"https://ror.org/008zs3103","country_code":"US","type":"education","lineage":["https://openalex.org/I74775410"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"T. S. Eugene Ng","raw_affiliation_strings":["Rice University, Houston, Texas, United States of America"],"raw_orcid":"https://orcid.org/0000-0003-2954-0767","affiliations":[{"raw_affiliation_string":"Rice University, Houston, Texas, United States of America","institution_ids":["https://openalex.org/I74775410"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":25,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"867","last_page":"882"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9995999932289124,"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":0.9995999932289124,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9962000250816345,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9961000084877014,"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/bottleneck","display_name":"Bottleneck","score":0.7383885979652405},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7299807667732239},{"id":"https://openalex.org/keywords/compression","display_name":"Compression (physics)","score":0.7077128887176514},{"id":"https://openalex.org/keywords/data-compression","display_name":"Data compression","score":0.6047527194023132},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5878562331199646},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.54768967628479},{"id":"https://openalex.org/keywords/tensor","display_name":"Tensor (intrinsic definition)","score":0.4848424196243286},{"id":"https://openalex.org/keywords/compression-ratio","display_name":"Compression ratio","score":0.430259108543396},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.430213987827301},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.40159887075424194},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.395637571811676},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.3077957332134247},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.17089316248893738},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10630470514297485},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.09353804588317871},{"id":"https://openalex.org/keywords/telecommunications","display_name":"Telecommunications","score":0.0874461829662323},{"id":"https://openalex.org/keywords/wireless","display_name":"Wireless","score":0.08588370680809021}],"concepts":[{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.7383885979652405},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7299807667732239},{"id":"https://openalex.org/C180016635","wikidata":"https://www.wikidata.org/wiki/Q2712821","display_name":"Compression (physics)","level":2,"score":0.7077128887176514},{"id":"https://openalex.org/C78548338","wikidata":"https://www.wikidata.org/wiki/Q2493","display_name":"Data compression","level":2,"score":0.6047527194023132},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5878562331199646},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.54768967628479},{"id":"https://openalex.org/C155281189","wikidata":"https://www.wikidata.org/wiki/Q3518150","display_name":"Tensor (intrinsic definition)","level":2,"score":0.4848424196243286},{"id":"https://openalex.org/C25797200","wikidata":"https://www.wikidata.org/wiki/Q828137","display_name":"Compression ratio","level":3,"score":0.430259108543396},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.430213987827301},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.40159887075424194},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.395637571811676},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3077957332134247},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.17089316248893738},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10630470514297485},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.09353804588317871},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0874461829662323},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.08588370680809021},{"id":"https://openalex.org/C511840579","wikidata":"https://www.wikidata.org/wiki/Q12757","display_name":"Internal combustion engine","level":2,"score":0.0},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0},{"id":"https://openalex.org/C192562407","wikidata":"https://www.wikidata.org/wiki/Q228736","display_name":"Materials science","level":0,"score":0.0},{"id":"https://openalex.org/C171146098","wikidata":"https://www.wikidata.org/wiki/Q124192","display_name":"Automotive engineering","level":1,"score":0.0},{"id":"https://openalex.org/C159985019","wikidata":"https://www.wikidata.org/wiki/Q181790","display_name":"Composite material","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3552326.3567505","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3552326.3567505","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Eighteenth European Conference on Computer Systems","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.5799999833106995,"display_name":"Decent work and economic growth","id":"https://metadata.un.org/sdg/8"}],"awards":[{"id":"https://openalex.org/G2064613317","display_name":null,"funder_award_id":"CNS-1815525","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G4662746506","display_name":null,"funder_award_id":"CNS-2214272","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":22,"referenced_works":["https://openalex.org/W2060393849","https://openalex.org/W2108598243","https://openalex.org/W2123442489","https://openalex.org/W2131613942","https://openalex.org/W2183341477","https://openalex.org/W2194775991","https://openalex.org/W2406293284","https://openalex.org/W2407022425","https://openalex.org/W2593390416","https://openalex.org/W2606722458","https://openalex.org/W2775776326","https://openalex.org/W2969388332","https://openalex.org/W2975712713","https://openalex.org/W2981114289","https://openalex.org/W2985108934","https://openalex.org/W2990190468","https://openalex.org/W3044837714","https://openalex.org/W3091097978","https://openalex.org/W3203134674","https://openalex.org/W3204921648","https://openalex.org/W4289401659","https://openalex.org/W6713184358"],"related_works":["https://openalex.org/W2891987081","https://openalex.org/W4383723869","https://openalex.org/W4384298135","https://openalex.org/W4383722264","https://openalex.org/W2388481516","https://openalex.org/W2161302774","https://openalex.org/W1723410974","https://openalex.org/W3007688875","https://openalex.org/W2152893502","https://openalex.org/W3209669196"],"abstract_inverted_index":{"Gradient":[0],"compression":[1,38,74,105,125],"(GC)":[2],"is":[3,34,68],"a":[4,123],"promising":[5],"approach":[6],"to":[7,42,70,79,102,120],"addressing":[8],"the":[9,19,37,46,57,72,83,91,108],"communication":[10,20,58],"bottleneck":[11],"in":[12],"distributed":[13],"deep":[14],"learning":[15],"(DDL).":[16],"It":[17],"saves":[18],"time,":[21],"but":[22],"also":[23],"incurs":[24],"additional":[25],"computation":[26],"overheads.":[27],"The":[28],"training":[29,116],"throughput":[30],"of":[31,48,82,93,113],"compression-enabled":[32],"DDL":[33,80,115],"determined":[35],"by":[36],"strategy,":[39],"including":[40],"whether":[41],"compress":[43],"each":[44],"tensor,":[45,62],"type":[47],"compute":[49],"resources":[50],"(e.g.,":[51],"CPUs":[52],"or":[53],"GPUs)":[54],"for":[55,60,76],"compression,":[56],"schemes":[59],"compressed":[61],"and":[63,107],"so":[64],"on.":[65],"However,":[66],"it":[67],"challenging":[69],"find":[71],"optimal":[73],"strategy":[75],"applying":[77],"GC":[78],"because":[81],"intricate":[84],"interactions":[85,110],"among":[86,111],"tensors.":[87],"To":[88],"fully":[89],"unleash":[90],"benefits":[92],"GC,":[94],"two":[95],"questions":[96],"must":[97],"be":[98],"addressed:":[99],"1)":[100],"How":[101,119],"express":[103],"any":[104,114],"strategies":[106],"corresponding":[109],"tensors":[112],"job?":[117],"2)":[118],"quickly":[121],"select":[122],"near-optimal":[124],"strategy?":[126]},"counts_by_year":[{"year":2026,"cited_by_count":4},{"year":2025,"cited_by_count":10},{"year":2024,"cited_by_count":10},{"year":2023,"cited_by_count":1}],"updated_date":"2026-07-22T07:51:19.307946","created_date":"2025-10-10T00:00:00"}