{"id":"https://openalex.org/W7166678938","doi":"https://doi.org/10.7148/2026-0638","title":"Benchmarking transformer-based time series models: pytorch vs mlx on apple silicon","display_name":"Benchmarking transformer-based time series models: pytorch vs mlx on apple silicon","publication_year":2026,"publication_date":"2026-06-23","ids":{"openalex":"https://openalex.org/W7166678938","doi":"https://doi.org/10.7148/2026-0638"},"language":null,"primary_location":{"id":"doi:10.7148/2026-0638","is_oa":true,"landing_page_url":"https://doi.org/10.7148/2026-0638","pdf_url":"https://doi.org/10.7148/2026-0638","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ECMS 2026 Proceedings edited by Filippo Sanfilippo, Florenc Demrozi, Fabio Sgarbossa, Mohammad Poursina","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.7148/2026-0638","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139679388","display_name":"Kamil Dziedzic","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kamil Dziedzic","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083798800","display_name":"Mateusz Nytko","orcid":"https://orcid.org/0000-0003-0606-1835"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mateusz Nytko","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139680516","display_name":"Filip Kruzel","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Filip Kruzel","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":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"638","last_page":"644"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12205","display_name":"Time Series Analysis and Forecasting","score":0.5372999906539917,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T12205","display_name":"Time Series Analysis and Forecasting","score":0.5372999906539917,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"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/T11512","display_name":"Anomaly Detection Techniques and Applications","score":0.2296999990940094,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T11326","display_name":"Stock Market Forecasting Methods","score":0.028999999165534973,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/benchmarking","display_name":"Benchmarking","score":0.6926000118255615},{"id":"https://openalex.org/keywords/series","display_name":"Series (stratigraphy)","score":0.5164999961853027},{"id":"https://openalex.org/keywords/hyperparameter","display_name":"Hyperparameter","score":0.4742000102996826},{"id":"https://openalex.org/keywords/anomaly-detection","display_name":"Anomaly detection","score":0.46720001101493835},{"id":"https://openalex.org/keywords/ranging","display_name":"Ranging","score":0.4523000121116638},{"id":"https://openalex.org/keywords/implementation","display_name":"Implementation","score":0.44110000133514404},{"id":"https://openalex.org/keywords/time-series","display_name":"Time series","score":0.4375},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.41280001401901245}],"concepts":[{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.6926000118255615},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6435999870300293},{"id":"https://openalex.org/C143724316","wikidata":"https://www.wikidata.org/wiki/Q312468","display_name":"Series (stratigraphy)","level":2,"score":0.5164999961853027},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4796000123023987},{"id":"https://openalex.org/C8642999","wikidata":"https://www.wikidata.org/wiki/Q4171168","display_name":"Hyperparameter","level":2,"score":0.4742000102996826},{"id":"https://openalex.org/C739882","wikidata":"https://www.wikidata.org/wiki/Q3560506","display_name":"Anomaly detection","level":2,"score":0.46720001101493835},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.46320000290870667},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.4523000121116638},{"id":"https://openalex.org/C26713055","wikidata":"https://www.wikidata.org/wiki/Q245962","display_name":"Implementation","level":2,"score":0.44110000133514404},{"id":"https://openalex.org/C151406439","wikidata":"https://www.wikidata.org/wiki/Q186588","display_name":"Time series","level":2,"score":0.4375},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.41280001401901245},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.4124999940395355},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.39579999446868896},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.3695000112056732},{"id":"https://openalex.org/C2780009758","wikidata":"https://www.wikidata.org/wiki/Q6804172","display_name":"Measure (data warehouse)","level":2,"score":0.35989999771118164},{"id":"https://openalex.org/C12997251","wikidata":"https://www.wikidata.org/wiki/Q567560","display_name":"Anomaly (physics)","level":2,"score":0.3474999964237213},{"id":"https://openalex.org/C139945424","wikidata":"https://www.wikidata.org/wiki/Q1940696","display_name":"Mean squared error","level":2,"score":0.31450000405311584},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.30230000615119934},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.29670000076293945},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.295199990272522},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2838999927043915},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.26489999890327454},{"id":"https://openalex.org/C3826847","wikidata":"https://www.wikidata.org/wiki/Q188768","display_name":"FLOPS","level":2,"score":0.2522999942302704},{"id":"https://openalex.org/C167085575","wikidata":"https://www.wikidata.org/wiki/Q6803654","display_name":"Mean squared prediction error","level":2,"score":0.25119999051094055},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.25}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.7148/2026-0638","is_oa":true,"landing_page_url":"https://doi.org/10.7148/2026-0638","pdf_url":"https://doi.org/10.7148/2026-0638","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ECMS 2026 Proceedings edited by Filippo Sanfilippo, Florenc Demrozi, Fabio Sgarbossa, Mohammad Poursina","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.7148/2026-0638","is_oa":true,"landing_page_url":"https://doi.org/10.7148/2026-0638","pdf_url":"https://doi.org/10.7148/2026-0638","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ECMS 2026 Proceedings edited by Filippo Sanfilippo, Florenc Demrozi, Fabio Sgarbossa, Mohammad Poursina","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.4051586985588074,"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure"}],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166678938.pdf","grobid_xml":"https://content.openalex.org/works/W7166678938.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"We":[0,157],"present":[1],"a":[2,50,139],"comparative":[3],"evaluation":[4],"of":[5],"three":[6],"transformer-based":[7],"time":[8,143],"series":[9,144],"models\u2014iTransformer,":[10],"PatchTST,":[11],"and":[12,22,42,49,65,105,127,168],"FEDformer\u2014across":[13],"two":[14],"deep":[15],"learning":[16],"frameworks:":[17],"the":[18,34,91,159],"original":[19],"PyTorch":[20,89],"implementations":[21],"newly":[23],"developed":[24],"ports":[25],"to":[26,100,162],"Apple\u2019s":[27],"MLX":[28,82,123,137],"framework.":[29],"Experiments":[30],"are":[31],"conducted":[32],"on":[33,90,147],"ETTm1":[35],"electricity":[36],"transformer":[37],"dataset,":[38],"covering":[39],"long-term":[40],"forecasting":[41],"anomaly":[43,77,121],"detection":[44],"tasks.":[45],"Using":[46],"identical":[47,131],"hyperparameters":[48],"consistent":[51],"80/20":[52],"chronological":[53],"train\u2013validation":[54],"split,":[55],"we":[56],"measure":[57],"training":[58,86,155],"throughput,":[59],"per-epoch":[60],"wall-clock":[61],"time,":[62],"memory":[63,166,180],"consumption,":[64],"predictive":[66],"accuracy":[67,109,152],"(RMSE,":[68],"MAE":[69],"for":[70,76,142],"forecasting;":[71],"precision,":[72],"recall,":[73],"F1,":[74],"AUC":[75,128],"detection).":[78],"Results":[79],"indicate":[80],"that":[81,136],"achieves":[83],"substantially":[84],"higher":[85,125],"throughput":[87],"than":[88],"M4":[92],"Pro":[93],"GPU,":[94],"with":[95,114,153,176],"speedups":[96,161],"ranging":[97],"from":[98],"1.02\u00d7":[99],"3.12\u00d7":[101],"across":[102],"model":[103,145],"architectures":[104],"batch":[106],"sizes.":[107],"Forecasting":[108],"remains":[110],"comparable":[111],"between":[112],"frameworks,":[113],"differences":[115,164],"in":[116,165],"RMSE":[117],"below":[118],"6%.":[119],"For":[120],"detection,":[122],"attains":[124],"F1":[126],"scores":[129],"under":[130],"configurations.":[132],"These":[133],"findings":[134],"suggest":[135],"is":[138],"viable":[140],"alternative":[141],"development":[146],"Apple":[148,177],"Silicon,":[149],"offering":[150],"competitive":[151],"improved":[154],"efficiency.":[156],"attribute":[158],"observed":[160],"architectural":[163],"handling":[167],"execution":[169],"models,":[170],"highlighting":[171],"how":[172],"framework":[173],"design":[174],"interacts":[175],"Silicon\u2019s":[178],"unified":[179],"architecture.":[181]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-07-01T00:00:00"}