{"id":"https://openalex.org/W4408151465","doi":"https://doi.org/10.1145/3658617.3697562","title":"End-to-end Compilation is All FPGAs Need: A Unified Overlay-based FPGA Compiler for Deep Learning","display_name":"End-to-end Compilation is All FPGAs Need: A Unified Overlay-based FPGA Compiler for Deep Learning","publication_year":2025,"publication_date":"2025-01-20","ids":{"openalex":"https://openalex.org/W4408151465","doi":"https://doi.org/10.1145/3658617.3697562"},"language":"en","primary_location":{"id":"doi:10.1145/3658617.3697562","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3658617.3697562","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th Asia and South Pacific Design Automation Conference","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/A5064508449","display_name":"Kai Qian","orcid":null},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kai Qian","raw_affiliation_strings":["Fudan Univ., Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-9522-9216","affiliations":[{"raw_affiliation_string":"Fudan Univ., Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047412399","display_name":"Haodong Lu","orcid":"https://orcid.org/0000-0002-8628-2664"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Haodong Lu","raw_affiliation_strings":["Fudan Univ., Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-8628-2664","affiliations":[{"raw_affiliation_string":"Fudan Univ., Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068961034","display_name":"Yinqiu Liu","orcid":"https://orcid.org/0000-0002-6198-3712"},"institutions":[{"id":"https://openalex.org/I172675005","display_name":"Nanyang Technological University","ror":"https://ror.org/02e7b5302","country_code":"SG","type":"education","lineage":["https://openalex.org/I172675005"]}],"countries":["SG"],"is_corresponding":false,"raw_author_name":"Yinqiu Liu","raw_affiliation_strings":["Nanyang Technological Univ., Singapore, Singapore"],"raw_orcid":"https://orcid.org/0000-0002-6198-3712","affiliations":[{"raw_affiliation_string":"Nanyang Technological Univ., Singapore, Singapore","institution_ids":["https://openalex.org/I172675005"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5043615810","display_name":"Zexu Zhang","orcid":"https://orcid.org/0009-0003-0931-2414"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zexu Zhang","raw_affiliation_strings":["Fudan Univ., Shanghai, China"],"raw_orcid":"https://orcid.org/0009-0003-0931-2414","affiliations":[{"raw_affiliation_string":"Fudan Univ., Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5015932760","display_name":"Kun Wang","orcid":"https://orcid.org/0000-0002-7288-1789"},"institutions":[{"id":"https://openalex.org/I24943067","display_name":"Fudan University","ror":"https://ror.org/013q1eq08","country_code":"CN","type":"education","lineage":["https://openalex.org/I24943067"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kun Wang","raw_affiliation_strings":["Fudan Univ., Shanghai, China"],"raw_orcid":"https://orcid.org/0000-0002-7288-1789","affiliations":[{"raw_affiliation_string":"Fudan Univ., Shanghai, China","institution_ids":["https://openalex.org/I24943067"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":3.0301,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.87301272,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":95,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"258","last_page":"264"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994999766349792,"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":0.9994999766349792,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.9988999962806702,"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":0.9975000023841858,"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/field-programmable-gate-array","display_name":"Field-programmable gate array","score":0.8286373019218445},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8178168535232544},{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.7922089695930481},{"id":"https://openalex.org/keywords/overlay","display_name":"Overlay","score":0.6854284405708313},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.4277108907699585},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.42734843492507935},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.41103607416152954},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.4022618532180786},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.2382224202156067},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.19521579146385193}],"concepts":[{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.8286373019218445},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8178168535232544},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.7922089695930481},{"id":"https://openalex.org/C136085584","wikidata":"https://www.wikidata.org/wiki/Q910289","display_name":"Overlay","level":2,"score":0.6854284405708313},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4277108907699585},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.42734843492507935},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.41103607416152954},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.4022618532180786},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.2382224202156067},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.19521579146385193}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3658617.3697562","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3658617.3697562","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 30th Asia and South Pacific Design Automation Conference","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":14,"referenced_works":["https://openalex.org/W2417429787","https://openalex.org/W2523838129","https://openalex.org/W2618530766","https://openalex.org/W2886241345","https://openalex.org/W2899644485","https://openalex.org/W2899691047","https://openalex.org/W2979455536","https://openalex.org/W3019166713","https://openalex.org/W3127141483","https://openalex.org/W3211485653","https://openalex.org/W4321637273","https://openalex.org/W4372258391","https://openalex.org/W4383749520","https://openalex.org/W4384947491"],"related_works":["https://openalex.org/W595346907","https://openalex.org/W153296825","https://openalex.org/W598989511","https://openalex.org/W2375779923","https://openalex.org/W2041986468","https://openalex.org/W1967800214","https://openalex.org/W2055675609","https://openalex.org/W1967938402","https://openalex.org/W2386041993","https://openalex.org/W1608572506"],"abstract_inverted_index":{"Field-Programmable":[0],"Gate":[1],"Array":[2],"(FPGA)":[3],"has":[4],"shown":[5],"great":[6],"application":[7],"potential":[8],"in":[9],"deploying":[10,25],"Neural":[11,118],"Networks":[12,119],"(NNs)":[13],"due":[14,41],"to":[15,42,66,90,97,114,176],"the":[16,51,59,104,111,147,153,183,187,200,206,211],"characteristics":[17],"of":[18,46,61,76,106,139,152,199],"programmability,":[19],"low":[20],"power":[21],"consumption,":[22],"etc.":[23],"However,":[24],"NNs":[26,34,89,142],"on":[27,137,160,219],"FPGA":[28,37,67,126,178,207],"is":[29],"non-trivial":[30],"because":[31],"(1)":[32,133],"Mainstream":[33],"pose":[35],"significant":[36,69],"architecture":[38],"design":[39],"challenges":[40],"their":[43],"large":[44,74],"number":[45],"parameters,":[47],"complex":[48],"operations,":[49],"and":[50,56,100,121,149,191,209,215,234,245],"need":[52],"for":[53,87,124,241],"data":[54,150,173],"optimization,":[55,194],"(2)":[57,167],"Supporting":[58],"deployment":[60],"different":[62,91,161],"machine":[63],"learning":[64],"frameworks":[65],"requires":[68],"manual":[70],"effort,":[71],"consuming":[72],"a":[73,84,140],"amount":[75],"time.":[77],"In":[78],"this":[79],"paper,":[80],"we":[81,109],"propose":[82],"AutoCompiler,":[83],"unified":[85],"compiler":[86,229],"mapping":[88],"FPGAs,":[92],"along":[93],"with":[94],"overlay":[95],"techniques":[96],"enable":[98],"fast":[99],"efficient":[101],"implementation.":[102],"To":[103],"best":[105],"our":[107],"knowledge,":[108],"are":[110],"first":[112],"work":[113],"support":[115],"both":[116],"Deep":[117],"(DNNs)":[120],"Transformer-based":[122],"networks":[123],"overlay-based":[125,228],"deployment.":[127],"AutoCompiler":[128,225],"comprises":[129],"three":[130],"integrated":[131],"enablers:":[132],"Model":[134],"Translator,":[135],"built":[136],"top":[138],"topology-based":[141],"representation,":[143],"which":[144,170,195],"can":[145],"optimize":[146],"topology":[148],"representation":[151],"models":[154],"from":[155],"an":[156],"algorithmic":[157],"level":[158,189],"based":[159],"hardware":[162],"configurations,":[163],"e.g.,":[164],"DSP":[165],"utilization,":[166],"Instruction":[168],"Generator,":[169],"generates":[171],"pipeline":[172],"streams":[174],"according":[175],"various":[177,220],"resource":[179],"configurations":[180],"by":[181,230,237],"manipulating":[182],"instruction":[184],"set":[185],"at":[186],"upper":[188],"rapidly,":[190],"(3)":[192],"End-to-end":[193],"moves":[196],"as":[197,203],"much":[198],"computational":[201],"processes":[202],"possible":[204],"onto":[205],"chip":[208],"minimizes":[210],"interaction":[212],"between":[213],"CPU":[214],"FPGA.":[216],"Extensive":[217],"experiments":[218],"Xilinx":[221],"FPGAs":[222],"show":[223],"that":[224],"outperforms":[226],"state-of-the-art":[227],"1.2\u00d7":[231],"-":[232,239],"1.35\u00d7":[233],"same-level":[235],"GPUs":[236],"1.15\u00d7":[238],"1.59\u00d7":[240],"classic":[242],"DNN":[243],"models,":[244],"ViT":[246],"inference,":[247],"respectively.":[248]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-10-10T00:00:00"}