{"id":"https://openalex.org/W2883830791","doi":"https://doi.org/10.1145/3302424.3303953","title":"Supporting Very Large Models using Automatic Dataflow Graph Partitioning","display_name":"Supporting Very Large Models using Automatic Dataflow Graph Partitioning","publication_year":2019,"publication_date":"2019-03-22","ids":{"openalex":"https://openalex.org/W2883830791","doi":"https://doi.org/10.1145/3302424.3303953","mag":"2883830791"},"language":"en","primary_location":{"id":"doi:10.1145/3302424.3303953","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3302424.3303953","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Fourteenth EuroSys Conference 2019","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/1807.08887","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101452217","display_name":"Minjie Wang","orcid":"https://orcid.org/0000-0001-8830-6079"},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Minjie Wang","raw_affiliation_strings":["New York University"],"affiliations":[{"raw_affiliation_string":"New York University","institution_ids":["https://openalex.org/I57206974"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109370237","display_name":"Chien-Chin Huang","orcid":null},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Chien-chin Huang","raw_affiliation_strings":["New York University"],"affiliations":[{"raw_affiliation_string":"New York University","institution_ids":["https://openalex.org/I57206974"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100320436","display_name":"Jinyang Li","orcid":"https://orcid.org/0009-0007-6936-4418"},"institutions":[{"id":"https://openalex.org/I57206974","display_name":"New York University","ror":"https://ror.org/0190ak572","country_code":"US","type":"education","lineage":["https://openalex.org/I57206974"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jinyang Li","raw_affiliation_strings":["New York University"],"affiliations":[{"raw_affiliation_string":"New York University","institution_ids":["https://openalex.org/I57206974"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5101452217"],"corresponding_institution_ids":["https://openalex.org/I57206974"],"apc_list":null,"apc_paid":null,"fwci":19.3049,"has_fulltext":false,"cited_by_count":137,"citation_normalized_percentile":{"value":0.99721166,"is_in_top_1_percent":true,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":90,"max":100},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"17"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9998000264167786,"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":0.9998000264167786,"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":0.9997000098228455,"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":0.9965000152587891,"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/dataflow","display_name":"Dataflow","score":0.9602565169334412},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8452945947647095},{"id":"https://openalex.org/keywords/memory-footprint","display_name":"Memory footprint","score":0.7370744347572327},{"id":"https://openalex.org/keywords/graph-partition","display_name":"Graph partition","score":0.6699384450912476},{"id":"https://openalex.org/keywords/partition","display_name":"Partition (number theory)","score":0.6524051427841187},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.6273723840713501},{"id":"https://openalex.org/keywords/graph","display_name":"Graph","score":0.5120156407356262},{"id":"https://openalex.org/keywords/operator","display_name":"Operator (biology)","score":0.440944105386734},{"id":"https://openalex.org/keywords/theoretical-computer-science","display_name":"Theoretical computer science","score":0.4387408196926117},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.42802396416664124},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.25301438570022583},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.08743077516555786}],"concepts":[{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.9602565169334412},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8452945947647095},{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.7370744347572327},{"id":"https://openalex.org/C48903430","wikidata":"https://www.wikidata.org/wiki/Q491370","display_name":"Graph partition","level":3,"score":0.6699384450912476},{"id":"https://openalex.org/C42812","wikidata":"https://www.wikidata.org/wiki/Q1082910","display_name":"Partition (number theory)","level":2,"score":0.6524051427841187},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.6273723840713501},{"id":"https://openalex.org/C132525143","wikidata":"https://www.wikidata.org/wiki/Q141488","display_name":"Graph","level":2,"score":0.5120156407356262},{"id":"https://openalex.org/C17020691","wikidata":"https://www.wikidata.org/wiki/Q139677","display_name":"Operator (biology)","level":5,"score":0.440944105386734},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.4387408196926117},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.42802396416664124},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.25301438570022583},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.08743077516555786},{"id":"https://openalex.org/C86339819","wikidata":"https://www.wikidata.org/wiki/Q407384","display_name":"Transcription factor","level":3,"score":0.0},{"id":"https://openalex.org/C104317684","wikidata":"https://www.wikidata.org/wiki/Q7187","display_name":"Gene","level":2,"score":0.0},{"id":"https://openalex.org/C55493867","wikidata":"https://www.wikidata.org/wiki/Q7094","display_name":"Biochemistry","level":1,"score":0.0},{"id":"https://openalex.org/C158448853","wikidata":"https://www.wikidata.org/wiki/Q425218","display_name":"Repressor","level":4,"score":0.0},{"id":"https://openalex.org/C114614502","wikidata":"https://www.wikidata.org/wiki/Q76592","display_name":"Combinatorics","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3302424.3303953","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3302424.3303953","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the Fourteenth EuroSys Conference 2019","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:1807.08887","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1807.08887","pdf_url":"https://arxiv.org/pdf/1807.08887","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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":"pmh:oai:arXiv.org:1807.08887","is_oa":true,"landing_page_url":"http://arxiv.org/abs/1807.08887","pdf_url":"https://arxiv.org/pdf/1807.08887","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"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"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G706268278","display_name":null,"funder_award_id":"CNS-1816717","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":107,"referenced_works":["https://openalex.org/W71081281","https://openalex.org/W78077100","https://openalex.org/W122068372","https://openalex.org/W141556878","https://openalex.org/W1442374986","https://openalex.org/W1484040084","https://openalex.org/W1498134579","https://openalex.org/W1522301498","https://openalex.org/W1569090332","https://openalex.org/W1598866093","https://openalex.org/W1598993252","https://openalex.org/W1635189588","https://openalex.org/W1724438581","https://openalex.org/W1833917188","https://openalex.org/W1845051632","https://openalex.org/W1939156876","https://openalex.org/W1947869163","https://openalex.org/W1964477602","https://openalex.org/W1969805974","https://openalex.org/W1973262089","https://openalex.org/W1980723535","https://openalex.org/W1987188645","https://openalex.org/W1997835767","https://openalex.org/W2006973305","https://openalex.org/W2008117760","https://openalex.org/W2019741404","https://openalex.org/W2024804689","https://openalex.org/W2040384644","https://openalex.org/W2053546108","https://openalex.org/W2053738729","https://openalex.org/W2055312318","https://openalex.org/W2056999868","https://openalex.org/W2064675550","https://openalex.org/W2083842231","https://openalex.org/W2090409324","https://openalex.org/W2090457779","https://openalex.org/W2093135480","https://openalex.org/W2093529653","https://openalex.org/W2102056800","https://openalex.org/W2109982223","https://openalex.org/W2112523971","https://openalex.org/W2115295617","https://openalex.org/W2124001993","https://openalex.org/W2127226421","https://openalex.org/W2130899092","https://openalex.org/W2130942839","https://openalex.org/W2140639562","https://openalex.org/W2146502635","https://openalex.org/W2153820245","https://openalex.org/W2159825340","https://openalex.org/W2162390675","https://openalex.org/W2168231600","https://openalex.org/W2173213060","https://openalex.org/W2186615578","https://openalex.org/W2188450683","https://openalex.org/W2194775991","https://openalex.org/W2259472270","https://openalex.org/W2261806027","https://openalex.org/W2267635276","https://openalex.org/W2281788661","https://openalex.org/W2289151794","https://openalex.org/W2335835108","https://openalex.org/W2338908902","https://openalex.org/W2339765813","https://openalex.org/W2340879669","https://openalex.org/W2402144811","https://openalex.org/W2423581336","https://openalex.org/W2433451779","https://openalex.org/W2518511512","https://openalex.org/W2525778437","https://openalex.org/W2553918877","https://openalex.org/W2557283755","https://openalex.org/W2559655401","https://openalex.org/W2602024037","https://openalex.org/W2605347906","https://openalex.org/W2786066748","https://openalex.org/W2786320458","https://openalex.org/W2786414509","https://openalex.org/W2798849118","https://openalex.org/W2804032941","https://openalex.org/W2883830791","https://openalex.org/W2884700152","https://openalex.org/W2911643614","https://openalex.org/W2949888546","https://openalex.org/W2952339051","https://openalex.org/W2952926545","https://openalex.org/W2962776038","https://openalex.org/W2963674932","https://openalex.org/W2963959650","https://openalex.org/W2964121744","https://openalex.org/W2964137095","https://openalex.org/W2964174152","https://openalex.org/W2964299589","https://openalex.org/W3092545530","https://openalex.org/W3101104221","https://openalex.org/W4205733209","https://openalex.org/W4239722617","https://openalex.org/W4240010000","https://openalex.org/W4251637954","https://openalex.org/W4299828299","https://openalex.org/W6679436768","https://openalex.org/W6683722107","https://openalex.org/W6693397755","https://openalex.org/W6713132643","https://openalex.org/W6713134421","https://openalex.org/W6730267373","https://openalex.org/W6748687944"],"related_works":["https://openalex.org/W2293118914","https://openalex.org/W2998381397","https://openalex.org/W2187181201","https://openalex.org/W4221139464","https://openalex.org/W4289827464","https://openalex.org/W3139829055","https://openalex.org/W2883830791","https://openalex.org/W2583811263","https://openalex.org/W3012931947","https://openalex.org/W2035944803"],"abstract_inverted_index":{"This":[0],"paper":[1],"presents":[2],"Tofu,":[3],"a":[4,26,57,69,74],"system":[5],"that":[6,78,91],"partitions":[7],"very":[8,97,115],"large":[9,98,116],"DNN":[10],"models":[11],"across":[12],"multiple":[13],"GPU":[14],"devices":[15],"to":[16,24,42,49,113],"reduce":[17],"per-GPU":[18],"memory":[19],"footprint.":[20],"Tofu":[21,72,92],"is":[22],"designed":[23],"partition":[25,44,65],"dataflow":[27,70],"graph":[28],"of":[29,53,96],"fine-grained":[30],"tensor":[31],"operators":[32,67],"used":[33],"by":[34,61],"platforms":[35],"like":[36],"MXNet":[37],"and":[38,100],"TensorFlow.":[39],"In":[40],"order":[41],"automatically":[43],"each":[45],"operator,":[46],"we":[47],"propose":[48],"describe":[50],"the":[51,80,94],"semantics":[52],"an":[54,87],"operator":[55],"in":[56,68],"simple":[58],"language":[59],"inspired":[60],"Halide.":[62],"To":[63],"optimally":[64],"different":[66],"graph,":[71],"uses":[73],"recursive":[75],"search":[76],"algorithm":[77],"minimizes":[79],"total":[81],"communication":[82],"cost.":[83],"Our":[84],"experiments":[85],"on":[86],"8-GPU":[88],"machine":[89],"show":[90],"enables":[93],"training":[95],"CNN":[99],"RNN":[101],"models.":[102,117],"It":[103],"also":[104],"achieves":[105],"25%":[106],"-":[107],"400%":[108],"speedup":[109],"over":[110],"alternative":[111],"approaches":[112],"train":[114]},"counts_by_year":[{"year":2026,"cited_by_count":3},{"year":2025,"cited_by_count":20},{"year":2024,"cited_by_count":24},{"year":2023,"cited_by_count":11},{"year":2022,"cited_by_count":19},{"year":2021,"cited_by_count":29},{"year":2020,"cited_by_count":19},{"year":2019,"cited_by_count":11},{"year":2018,"cited_by_count":1}],"updated_date":"2026-04-02T15:55:50.835912","created_date":"2025-10-10T00:00:00"}