{"id":"https://openalex.org/W4406825384","doi":"https://doi.org/10.1145/3715122","title":"Automatic Generation of Fast and Accurate Performance Models for Deep Neural Network Accelerators","display_name":"Automatic Generation of Fast and Accurate Performance Models for Deep Neural Network Accelerators","publication_year":2025,"publication_date":"2025-01-25","ids":{"openalex":"https://openalex.org/W4406825384","doi":"https://doi.org/10.1145/3715122"},"language":"en","primary_location":{"id":"doi:10.1145/3715122","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3715122","pdf_url":null,"source":{"id":"https://openalex.org/S136160450","display_name":"ACM Transactions on Embedded Computing Systems","issn_l":"1539-9087","issn":["1539-9087","1558-3465"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Embedded Computing Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1145/3715122","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5069328202","display_name":"Konstantin L\u00fcbeck","orcid":"https://orcid.org/0000-0002-2701-5881"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Konstantin L\u00fcbeck","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0002-2701-5881","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073430594","display_name":"A. Jung","orcid":"https://orcid.org/0000-0001-5702-5768"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Alexander Louis-Ferdinand Jung","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0001-5702-5768","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5045329350","display_name":"Felix Wedlich","orcid":"https://orcid.org/0009-0007-2632-7078"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Felix Wedlich","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0009-0007-2632-7078","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111123870","display_name":"Mika Markus M\u00fcller","orcid":"https://orcid.org/0009-0003-3471-7560"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Mika Markus M\u00fcller","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0009-0003-3471-7560","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5021702622","display_name":"Federico Nicol\u00e1s Peccia","orcid":"https://orcid.org/0000-0002-3587-0415"},"institutions":[{"id":"https://openalex.org/I143379178","display_name":"FZI Research Center for Information Technology","ror":"https://ror.org/04kdh6x72","country_code":"DE","type":"nonprofit","lineage":["https://openalex.org/I143379178"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Federico Nicol\u00e1s Peccia","raw_affiliation_strings":["FZI, Karlsruhe, Germany","FZI,  Karlsruhe, Germany"],"raw_orcid":"https://orcid.org/0000-0002-3587-0415","affiliations":[{"raw_affiliation_string":"FZI, Karlsruhe, Germany","institution_ids":["https://openalex.org/I143379178"]},{"raw_affiliation_string":"FZI,  Karlsruhe, Germany","institution_ids":["https://openalex.org/I143379178"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5093577302","display_name":"Felix Th\u00f6mmes","orcid":"https://orcid.org/0009-0007-0885-5424"},"institutions":[{"id":"https://openalex.org/I143379178","display_name":"FZI Research Center for Information Technology","ror":"https://ror.org/04kdh6x72","country_code":"DE","type":"nonprofit","lineage":["https://openalex.org/I143379178"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Felix Th\u00f6mmes","raw_affiliation_strings":["FZI, Karlsruhe, Germany","FZI,  Karlsruhe, Germany"],"raw_orcid":"https://orcid.org/0009-0007-0885-5424","affiliations":[{"raw_affiliation_string":"FZI, Karlsruhe, Germany","institution_ids":["https://openalex.org/I143379178"]},{"raw_affiliation_string":"FZI,  Karlsruhe, Germany","institution_ids":["https://openalex.org/I143379178"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5099124256","display_name":"Jannik Steinmetz","orcid":"https://orcid.org/0009-0003-7193-5620"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Jannik Steinmetz","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0009-0003-7193-5620","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114367781","display_name":"Valentin Biermaier","orcid":"https://orcid.org/0009-0007-9428-9085"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Valentin Biermaier","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0009-0007-9428-9085","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5041394449","display_name":"Adrian Frischknecht","orcid":"https://orcid.org/0000-0002-1795-0948"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Adrian Frischknecht","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0002-1795-0948","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5007898502","display_name":"Paul Palomero Bernardo","orcid":"https://orcid.org/0000-0002-6642-3976"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Paul Palomero Bernardo","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0002-6642-3976","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5074802358","display_name":"Oliver Bringmann","orcid":"https://orcid.org/0000-0002-1615-507X"},"institutions":[{"id":"https://openalex.org/I8087733","display_name":"University of T\u00fcbingen","ror":"https://ror.org/03a1kwz48","country_code":"DE","type":"education","lineage":["https://openalex.org/I8087733"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Oliver Bringmann","raw_affiliation_strings":["Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany"],"raw_orcid":"https://orcid.org/0000-0002-1615-507X","affiliations":[{"raw_affiliation_string":"Embedded Systems, University of T\u00fcbingen, T\u00fcbingen, Germany","institution_ids":["https://openalex.org/I8087733"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.00878638,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"24","issue":"2","first_page":"1","last_page":"32"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9994000196456909,"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":0.9994000196456909,"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":0.9984999895095825,"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/T11992","display_name":"CCD and CMOS Imaging Sensors","score":0.9905999898910522,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.6310524940490723},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.5966454744338989},{"id":"https://openalex.org/keywords/deep-neural-networks","display_name":"Deep neural networks","score":0.46393969655036926},{"id":"https://openalex.org/keywords/deep-learning","display_name":"Deep learning","score":0.46366196870803833},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.4611279368400574}],"concepts":[{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.6310524940490723},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5966454744338989},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.46393969655036926},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.46366196870803833},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4611279368400574}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3715122","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3715122","pdf_url":null,"source":{"id":"https://openalex.org/S136160450","display_name":"ACM Transactions on Embedded Computing Systems","issn_l":"1539-9087","issn":["1539-9087","1558-3465"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Embedded Computing Systems","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1145/3715122","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3715122","pdf_url":null,"source":{"id":"https://openalex.org/S136160450","display_name":"ACM Transactions on Embedded Computing Systems","issn_l":"1539-9087","issn":["1539-9087","1558-3465"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Transactions on Embedded Computing Systems","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321114","display_name":"Bundesministerium f\u00fcr Bildung und Forschung","ror":"https://ror.org/04pz7b180"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":27,"referenced_works":["https://openalex.org/W1983394510","https://openalex.org/W2002555321","https://openalex.org/W2064639033","https://openalex.org/W2171074980","https://openalex.org/W2289252105","https://openalex.org/W2431931973","https://openalex.org/W2618530766","https://openalex.org/W2626211758","https://openalex.org/W2940862705","https://openalex.org/W2949251082","https://openalex.org/W2973226577","https://openalex.org/W2995763407","https://openalex.org/W2998732502","https://openalex.org/W3027968530","https://openalex.org/W3092482740","https://openalex.org/W3098316065","https://openalex.org/W3113750219","https://openalex.org/W3157708976","https://openalex.org/W3213528054","https://openalex.org/W4210482024","https://openalex.org/W4231250608","https://openalex.org/W4247199268","https://openalex.org/W4253012315","https://openalex.org/W4280600300","https://openalex.org/W4308071528","https://openalex.org/W4313496578","https://openalex.org/W4391514505"],"related_works":["https://openalex.org/W2731899572","https://openalex.org/W3215138031","https://openalex.org/W3009238340","https://openalex.org/W4360585206","https://openalex.org/W4321369474","https://openalex.org/W4285208911","https://openalex.org/W3082895349","https://openalex.org/W4213079790","https://openalex.org/W2248239756","https://openalex.org/W3086377361"],"abstract_inverted_index":{"Implementing":[0],"Deep":[1],"Neural":[2],"Networks":[3],"(DNNs)":[4],"on":[5],"resource-constrained":[6],"edge":[7],"devices":[8],"is":[9],"a":[10,20,52,80,94,123],"challenging":[11],"task":[12],"that":[13],"requires":[14],"tailored":[15],"hardware":[16],"accelerator":[17,61,65],"architectures":[18],"and":[19,58,79,129],"clear":[21],"understanding":[22],"of":[23,51,134],"their":[24],"performance":[25,44,117],"characteristics":[26],"when":[27],"executing":[28],"the":[29,49,104,116],"intended":[30],"AI":[31],"workload.":[32],"To":[33],"facilitate":[34],"this,":[35],"we":[36,69,92],"present":[37],"an":[38,150],"automated":[39],"generation":[40],"approach":[41],"for":[42,88,118],"fast":[43],"models":[45,131],"to":[46,107,114],"accurately":[47],"estimate":[48,115],"latency":[50],"DNN":[53,72,86],"mapped":[54],"onto":[55],"systematically":[56],"modeled":[57,70,90],"concisely":[59],"described":[60],"architectures.":[62],"Using":[63],"our":[64],"architecture":[66],"description":[67],"method,":[68],"representative":[71],"accelerators":[73],"such":[74],"as":[75],"Gemmini,":[76],"UltraTrail,":[77],"Plasticine-derived,":[78],"parameterizable":[81],"systolic":[82],"array.":[83],"Together":[84],"with":[85,141],"mappings":[87],"those":[89],"architectures,":[91],"perform":[93],"combined":[95],"DNN/hardware":[96],"dependency":[97],"graph":[98],"analysis,":[99],"which":[100],"enables":[101],"us,":[102],"in":[103,132],"best":[105],"case,":[106],"evaluate":[108],"only":[109],"154":[110],"loop":[111],"kernel":[112],"iterations":[113],"4.19":[119],"billion":[120],"instructions":[121],"achieving":[122],"significant":[124],"speedup.":[125],"We":[126],"outperform":[127],"regression":[128],"analytical":[130],"terms":[133],"mean":[135],"absolute":[136],"percentage":[137],"error":[138],"(MAPE)":[139],"compared":[140],"simulation":[142],"results,":[143],"while":[144],"being":[145],"several":[146],"magnitudes":[147],"faster":[148],"than":[149],"RTL":[151],"simulation.":[152]},"counts_by_year":[],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2025-01-26T00:00:00"}