{"id":"https://openalex.org/W7165639079","doi":"https://doi.org/10.48550/arxiv.2606.21101","title":"DPIFrame: A Dual-Level Parallelism Acceleration Framework for CTR Model Inference","display_name":"DPIFrame: A Dual-Level Parallelism Acceleration Framework for CTR Model Inference","publication_year":2026,"publication_date":"2026-06-19","ids":{"openalex":"https://openalex.org/W7165639079","doi":"https://doi.org/10.48550/arxiv.2606.21101"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.21101","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21101","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.21101","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5092966843","display_name":"Dezhi Yi","orcid":"https://orcid.org/0009-0004-3553-734X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yi, Dezhi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019541869","display_name":"Huifeng Guo","orcid":"https://orcid.org/0000-0002-7393-8994"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Guo, Huifeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102770065","display_name":"Kunpeng Xie","orcid":"https://orcid.org/0000-0001-9857-5352"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Kunpeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071690387","display_name":"Zhaolong Jian","orcid":"https://orcid.org/0000-0003-1543-3207"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jian, Zhaolong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030802038","display_name":"Haochi Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Haochi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069292806","display_name":"Wenxuan He","orcid":"https://orcid.org/0000-0001-6979-4564"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Wenxuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139184701","display_name":"Zhenhua Dong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, Zhenhua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139211260","display_name":"Ruiming Tang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tang, Ruiming","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139153670","display_name":"Ye Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Ye","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.7095000147819519,"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":0.7095000147819519,"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":0.023099999874830246,"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/T14347","display_name":"Big Data and Digital Economy","score":0.018799999728798866,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/parallelizable-manifold","display_name":"Parallelizable manifold","score":0.8172000050544739},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6506999731063843},{"id":"https://openalex.org/keywords/embedding","display_name":"Embedding","score":0.602400004863739},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.5529999732971191},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.4659999907016754},{"id":"https://openalex.org/keywords/approximate-inference","display_name":"Approximate inference","score":0.4291999936103821},{"id":"https://openalex.org/keywords/workload","display_name":"Workload","score":0.3529999852180481},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.3476000130176544},{"id":"https://openalex.org/keywords/parallelism","display_name":"Parallelism (grammar)","score":0.34130001068115234}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8583999872207642},{"id":"https://openalex.org/C148047603","wikidata":"https://www.wikidata.org/wiki/Q1014612","display_name":"Parallelizable manifold","level":2,"score":0.8172000050544739},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6506999731063843},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.6136999726295471},{"id":"https://openalex.org/C41608201","wikidata":"https://www.wikidata.org/wiki/Q980509","display_name":"Embedding","level":2,"score":0.602400004863739},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5529999732971191},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.4659999907016754},{"id":"https://openalex.org/C2777472644","wikidata":"https://www.wikidata.org/wiki/Q16968992","display_name":"Approximate inference","level":3,"score":0.4291999936103821},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.3529999852180481},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.3476000130176544},{"id":"https://openalex.org/C2781172179","wikidata":"https://www.wikidata.org/wiki/Q853109","display_name":"Parallelism (grammar)","level":2,"score":0.34130001068115234},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3384999930858612},{"id":"https://openalex.org/C107027933","wikidata":"https://www.wikidata.org/wiki/Q2006448","display_name":"Stream processing","level":2,"score":0.33219999074935913},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.32910001277923584},{"id":"https://openalex.org/C120373497","wikidata":"https://www.wikidata.org/wiki/Q1087987","display_name":"Parallel algorithm","level":2,"score":0.3222000002861023},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.3140000104904175},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.30059999227523804},{"id":"https://openalex.org/C139002025","wikidata":"https://www.wikidata.org/wiki/Q3001212","display_name":"Lift (data mining)","level":2,"score":0.29010000824928284},{"id":"https://openalex.org/C157170001","wikidata":"https://www.wikidata.org/wiki/Q4781507","display_name":"Applications of artificial intelligence","level":2,"score":0.28780001401901245},{"id":"https://openalex.org/C117896860","wikidata":"https://www.wikidata.org/wiki/Q11376","display_name":"Acceleration","level":2,"score":0.2777000069618225},{"id":"https://openalex.org/C201410400","wikidata":"https://www.wikidata.org/wiki/Q1064412","display_name":"Multithreading","level":3,"score":0.27559998631477356},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.275299996137619},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2669999897480011},{"id":"https://openalex.org/C106515295","wikidata":"https://www.wikidata.org/wiki/Q26806595","display_name":"Parallel processing","level":2,"score":0.2651999890804291},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.25690001249313354},{"id":"https://openalex.org/C61483411","wikidata":"https://www.wikidata.org/wiki/Q3124522","display_name":"Data parallelism","level":3,"score":0.25600001215934753},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.25529998540878296},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.25189998745918274},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2502000033855438}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.21101","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21101","pdf_url":null,"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":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.21101","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21101","pdf_url":null,"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":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Deep":[0],"learning":[1],"technology":[2],"has":[3],"enhanced":[4],"the":[5,42,47,57,100,124,137,144],"ability":[6],"of":[7,116,170],"Click-through":[8],"rate":[9],"(CTR)":[10],"prediction":[11,18],"models":[12,27],"to":[13,24,62,76,121,151],"learn":[14],"features":[15],"and":[16,31,46,85,136,158,174],"improve":[17],"accuracy.":[19],"However,":[20],"it":[21],"is":[22,37,74,93,111],"challenging":[23],"deploy":[25],"CTR":[26,64,79],"on":[28,119,132,166],"GPU":[29,120,167],"smoothly":[30],"perform":[32,77],"inference":[33,81,164],"efficiently,":[34],"because":[35],"there":[36],"a":[38,70,106],"huge":[39],"mismatch":[40],"between":[41],"serial":[43],"computational":[44],"pattern":[45],"parallel":[48,78,117,126],"model":[49,65,80],"structure.":[50],"In":[51,67],"this":[52],"paper,":[53],"we":[54],"propose":[55],"DPIFrame,":[56,68],"first":[58],"dual":[59,71,125],"parallelizable":[60,72],"framework":[61],"accelerate":[63],"inference.":[66],"a)":[69],"architecture":[73],"proposed":[75],"in":[82,103],"both":[83],"intra-module":[84],"inter-module;":[86],"b)":[87],"an":[88],"efficient":[89],"multi-table":[90],"lookup":[91],"algorithm":[92],"presented":[94],"for":[95,113],"embedding":[96,145],"operations":[97],"through":[98],"anticipating":[99],"whole":[101],"workload":[102],"advance;":[104],"c)":[105],"breadth-first":[107],"stream":[108],"scheduling":[109],"strategy":[110],"designed":[112],"fine-grained":[114],"management":[115],"computation":[118],"further":[122],"supporting":[123],"execution.":[127],"Extensive":[128],"experiments":[129],"are":[130],"conducted":[131],"two":[133],"real-world":[134],"datasets,":[135],"results":[138],"highlight":[139],"that":[140],"DPIFrame":[141,160],"can":[142,161],"reduce":[143],"latency":[146],"efficiently":[147],"by":[148],"\\textbf{23.0$\\times$}":[149],"compared":[150],"PyTorch.":[152],"Compared":[153],"with":[154,168],"PyTorch,":[155],"TorchRec,":[156],"HugeCTR,":[157],"OneFlow,":[159],"achieve":[162],"state-of-the-art":[163],"performance":[165],"speedups":[169],"\\textbf{5.83$\\times$},":[171],"\\textbf{4.29$\\times$},":[172],"\\textbf{2.15$\\times$},":[173],"\\textbf{2.0$\\times$},":[175],"respectively.":[176]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-24T00:00:00"}