{"id":"https://openalex.org/W7131665847","doi":"https://doi.org/10.1007/s11227-026-08327-6","title":"Enhancing transformer performance and portability through auto-tuning frameworks","display_name":"Enhancing transformer performance and portability through auto-tuning frameworks","publication_year":2026,"publication_date":"2026-02-26","ids":{"openalex":"https://openalex.org/W7131665847","doi":"https://doi.org/10.1007/s11227-026-08327-6"},"language":"en","primary_location":{"id":"doi:10.1007/s11227-026-08327-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-026-08327-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-026-08327-6.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://link.springer.com/content/pdf/10.1007/s11227-026-08327-6.pdf","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120169657","display_name":"Patricia Siwinska","orcid":null},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Patricia Siwinska","raw_affiliation_strings":["Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain","institution_ids":["https://openalex.org/I60053951"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126941276","display_name":"Jie Lei","orcid":null},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Jie Lei","raw_affiliation_strings":["Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain","institution_ids":["https://openalex.org/I60053951"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123587466","display_name":"Adri\u00e1n Castell\u00f3","orcid":null},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]}],"countries":["ES"],"is_corresponding":true,"raw_author_name":"Adri\u00e1n Castell\u00f3","raw_affiliation_strings":["Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain","institution_ids":["https://openalex.org/I60053951"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126931581","display_name":"Pedro Alonso-Jord\u00e1","orcid":null},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Pedro Alonso-Jord\u00e1","raw_affiliation_strings":["Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain","institution_ids":["https://openalex.org/I60053951"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5126901945","display_name":"Enrique S. Quintana-Ort\u00ed","orcid":null},"institutions":[{"id":"https://openalex.org/I60053951","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859","country_code":"ES","type":"education","lineage":["https://openalex.org/I60053951"]}],"countries":["ES"],"is_corresponding":false,"raw_author_name":"Enrique S. Quintana-Ort\u00ed","raw_affiliation_strings":["Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Universitat Polit\u00e8cnica de Val\u00e8ncia, Val\u00e8ncia, Spain","institution_ids":["https://openalex.org/I60053951"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":["https://openalex.org/A5123587466"],"corresponding_institution_ids":["https://openalex.org/I60053951"],"apc_list":{"value":2390,"currency":"EUR","value_usd":2990},"apc_paid":{"value":2390,"currency":"EUR","value_usd":2990},"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.14826192,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"82","issue":"4","first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.820900022983551,"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":0.820900022983551,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.029100000858306885,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T11697","display_name":"Numerical Methods and Algorithms","score":0.014700000174343586,"subfield":{"id":"https://openalex.org/subfields/1703","display_name":"Computational Theory and Mathematics"},"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/software-portability","display_name":"Software portability","score":0.8087000250816345},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.6513000130653381},{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.6438999772071838},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.5396999716758728},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.44029998779296875},{"id":"https://openalex.org/keywords/implementation","display_name":"Implementation","score":0.4397999942302704},{"id":"https://openalex.org/keywords/programming-paradigm","display_name":"Programming paradigm","score":0.41530001163482666},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.39100000262260437},{"id":"https://openalex.org/keywords/x86","display_name":"x86","score":0.3549000024795532}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.885699987411499},{"id":"https://openalex.org/C63000827","wikidata":"https://www.wikidata.org/wiki/Q3080428","display_name":"Software portability","level":2,"score":0.8087000250816345},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.6513000130653381},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.6438999772071838},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.5396999716758728},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.44029998779296875},{"id":"https://openalex.org/C26713055","wikidata":"https://www.wikidata.org/wiki/Q245962","display_name":"Implementation","level":2,"score":0.4397999942302704},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.4341000020503998},{"id":"https://openalex.org/C34165917","wikidata":"https://www.wikidata.org/wiki/Q188267","display_name":"Programming paradigm","level":2,"score":0.41530001163482666},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.41449999809265137},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.39100000262260437},{"id":"https://openalex.org/C170723468","wikidata":"https://www.wikidata.org/wiki/Q182933","display_name":"x86","level":3,"score":0.3549000024795532},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3467000126838684},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.33809998631477356},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.3287000060081482},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.32749998569488525},{"id":"https://openalex.org/C76782552","wikidata":"https://www.wikidata.org/wiki/Q110546","display_name":"Just-in-time compilation","level":3,"score":0.31470000743865967},{"id":"https://openalex.org/C2982832238","wikidata":"https://www.wikidata.org/wiki/Q5531640","display_name":"General purpose","level":2,"score":0.3066999912261963},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.3037000000476837},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.2992999851703644},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.2919999957084656},{"id":"https://openalex.org/C2780870223","wikidata":"https://www.wikidata.org/wiki/Q1004415","display_name":"Runtime system","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.28600001335144043},{"id":"https://openalex.org/C2778565505","wikidata":"https://www.wikidata.org/wiki/Q2207566","display_name":"Spec#","level":2,"score":0.26899999380111694},{"id":"https://openalex.org/C190902152","wikidata":"https://www.wikidata.org/wiki/Q1325106","display_name":"Optimizing compiler","level":3,"score":0.25850000977516174},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.25839999318122864},{"id":"https://openalex.org/C85687889","wikidata":"https://www.wikidata.org/wiki/Q445962","display_name":"Legacy code","level":3,"score":0.25519999861717224},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.25200000405311584},{"id":"https://openalex.org/C45357846","wikidata":"https://www.wikidata.org/wiki/Q2001982","display_name":"Notation","level":2,"score":0.25099998712539673}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1007/s11227-026-08327-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-026-08327-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-026-08327-6.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1007/s11227-026-08327-6","is_oa":true,"landing_page_url":"https://doi.org/10.1007/s11227-026-08327-6","pdf_url":"https://link.springer.com/content/pdf/10.1007/s11227-026-08327-6.pdf","source":{"id":"https://openalex.org/S32326811","display_name":"The Journal of Supercomputing","issn_l":"0920-8542","issn":["0920-8542","1573-0484"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319900","host_organization_name":"Springer Science+Business Media","host_organization_lineage":["https://openalex.org/P4310319900","https://openalex.org/P4310319965"],"host_organization_lineage_names":["Springer Science+Business Media","Springer Nature"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"The Journal of Supercomputing","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G141703365","display_name":null,"funder_award_id":"PROMETEO 2023-CIPROM/2022/20","funder_id":"https://openalex.org/F4320321864","funder_display_name":"Generalitat Valenciana"},{"id":"https://openalex.org/G1422035950","display_name":null,"funder_award_id":"501100011033","funder_id":"https://openalex.org/F4320321864","funder_display_name":"Generalitat Valenciana"},{"id":"https://openalex.org/G2262748287","display_name":null,"funder_award_id":"501100011033","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G2867640486","display_name":null,"funder_award_id":"13039/501100011033","funder_id":"https://openalex.org/F4320334905","funder_display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia"},{"id":"https://openalex.org/G3010678871","display_name":null,"funder_award_id":"PAID_06_24","funder_id":"https://openalex.org/F4320334905","funder_display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia"},{"id":"https://openalex.org/G3480869486","display_name":null,"funder_award_id":"13039","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G5005885202","display_name":null,"funder_award_id":"PID2023-146569NB-C2","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G7084143925","display_name":null,"funder_award_id":"AEI/10","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G7266728691","display_name":null,"funder_award_id":"13039/501100011033","funder_id":"https://openalex.org/F4320335598","funder_display_name":"Agencia Estatal de Investigaci\u00f3n"},{"id":"https://openalex.org/G7268358848","display_name":null,"funder_award_id":"CIPROM/2022/20","funder_id":"https://openalex.org/F4320321864","funder_display_name":"Generalitat Valenciana"}],"funders":[{"id":"https://openalex.org/F4320321864","display_name":"Generalitat Valenciana","ror":"https://ror.org/0097mvx21"},{"id":"https://openalex.org/F4320334905","display_name":"Universitat Polit\u00e8cnica de Val\u00e8ncia","ror":"https://ror.org/01460j859"},{"id":"https://openalex.org/F4320335598","display_name":"Agencia Estatal de Investigaci\u00f3n","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7131665847.pdf","grobid_xml":"https://content.openalex.org/works/W7131665847.grobid-xml"},"referenced_works_count":11,"referenced_works":["https://openalex.org/W2000050212","https://openalex.org/W2026049208","https://openalex.org/W2055312318","https://openalex.org/W2094969361","https://openalex.org/W2899585750","https://openalex.org/W4281707342","https://openalex.org/W4390224392","https://openalex.org/W4392265981","https://openalex.org/W4404742986","https://openalex.org/W4409427297","https://openalex.org/W4412397574"],"related_works":[],"abstract_inverted_index":{"Abstract":[0],"Transformer-based":[1],"models":[2],"such":[3],"as":[4,52,86,174],"BERT":[5],"and":[6,22,34,50,69,78,97,102,104,112,126,129,136,142,169,177],"GPT2":[7],"have":[8,37],"become":[9],"the":[10,80,93,153,165],"foundation":[11],"of":[12,82,92,167],"many":[13],"modern":[14],"applications,":[15],"yet":[16],"their":[17,106],"execution":[18,73],"requires":[19],"substantial":[20],"computational":[21],"memory":[23],"resources.":[24],"To":[25],"address":[26],"these":[27],"challenges,":[28],"recent":[29],"advances":[30],"in":[31,79,100,184],"compiler":[32],"technology":[33],"hardware":[35,120],"accelerators":[36],"introduced":[38],"new":[39],"opportunities":[40],"for":[41,180],"performance":[42,146],"portability.":[43],"In":[44],"this":[45],"work,":[46],"we":[47],"evaluate":[48],"JAX":[49,101,141],"TVM":[51,103,143],"high-level":[53],"frameworks":[54,171],"that":[55,140],"combine":[56],"a":[57,175],"NumPy-like":[58],"programming":[59,154],"model":[60],"with":[61],"Just-In-Time":[62],"(JIT)":[63],"or":[64,76],"Ahead-of-Time":[65],"(AOT)":[66],"code":[67,109],"optimization":[68],"compilation,":[70],"enabling":[71],"efficient":[72,178],"across":[74],"CPUs":[75],"GPUs,":[77,125],"case":[81],"JAX,":[83],"on":[84],"TPUs":[85],"well.":[87],"We":[88],"present":[89],"systematic":[90],"implementations":[91],"core":[94],"Transformer":[95,182],"encoder":[96],"decoder":[98],"blocks":[99],"compare":[105],"automatically":[107],"optimized":[108],"against":[110],"NumPy":[111],"CuPy":[113],"baselines.":[114],"Our":[115],"experimental":[116],"study":[117],"covers":[118],"heterogeneous":[119],"platforms":[121],"(AMD":[122],"CPU,":[123],"NVIDIA":[124],"Google":[127],"TPUs)":[128],"multiple":[130],"arithmetic":[131],"precisions":[132],"(FP32,":[133],"BF16,":[134],"INT8,":[135],"INT32).":[137],"Results":[138],"show":[139],"deliver":[144],"significant":[145],"improvements":[147],"over":[148],"standard":[149],"libraries,":[150],"while":[151],"reducing":[152],"effort":[155],"required":[156],"to":[157,159,172],"adapt":[158],"different":[160],"hardware.":[161],"These":[162],"findings":[163],"demonstrate":[164],"potential":[166],"JIT-":[168],"AOT-oriented":[170],"serve":[173],"portable":[176],"solution":[179],"deploying":[181],"workloads":[183],"diverse":[185],"computing":[186],"environments.":[187]},"counts_by_year":[],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2026-02-27T00:00:00"}
