{"id":"https://openalex.org/W4409248462","doi":"https://doi.org/10.1109/hpca61900.2025.00128","title":"Lincoln: Real-Time 50~100B LLM Inference on Consumer Devices with LPDDR-Interfaced, Compute-Enabled Flash Memory","display_name":"Lincoln: Real-Time 50~100B LLM Inference on Consumer Devices with LPDDR-Interfaced, Compute-Enabled Flash Memory","publication_year":2025,"publication_date":"2025-03-01","ids":{"openalex":"https://openalex.org/W4409248462","doi":"https://doi.org/10.1109/hpca61900.2025.00128"},"language":"en","primary_location":{"id":"doi:10.1109/hpca61900.2025.00128","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00128","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"},"type":"article","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/A5101979303","display_name":"Weiyi Sun","orcid":"https://orcid.org/0000-0002-0432-6636"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Weiyi Sun","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5026297396","display_name":"Mingyu Gao","orcid":"https://orcid.org/0000-0001-8433-7281"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Mingyu Gao","raw_affiliation_strings":["Institute for Interdisciplinary Information Sciences, Tsinghua University, Shanghai Qi Zhi Institute,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Institute for Interdisciplinary Information Sciences, Tsinghua University, Shanghai Qi Zhi Institute,Shanghai,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5070301498","display_name":"Zhaoshi Li","orcid":"https://orcid.org/0000-0003-0786-9350"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhaoshi Li","raw_affiliation_strings":["MetaX Technology,Beijing,China"],"affiliations":[{"raw_affiliation_string":"MetaX Technology,Beijing,China","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5083014834","display_name":"Aoyang Zhang","orcid":"https://orcid.org/0000-0002-2128-5274"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Aoyang Zhang","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5117071229","display_name":"Iris Ying Chou","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Iris Ying Chou","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5102891205","display_name":"Jianfeng Zhu","orcid":"https://orcid.org/0000-0002-0485-8034"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianfeng Zhu","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032344424","display_name":"Shaojun Wei","orcid":null},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shaojun Wei","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100358856","display_name":"Leibo Liu","orcid":"https://orcid.org/0000-0001-7548-4116"},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Leibo Liu","raw_affiliation_strings":["Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China"],"affiliations":[{"raw_affiliation_string":"Tsinghua University, Beijing National Research Center for Information Science and Technology (BNRist),School of Integrated Circuits,Beijing,China","institution_ids":["https://openalex.org/I99065089"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5101979303"],"corresponding_institution_ids":["https://openalex.org/I99065089"],"apc_list":null,"apc_paid":null,"fwci":8.0118,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.97076866,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":96,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"1734","last_page":"1750"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.3068999946117401,"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"}},"topics":[{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.3068999946117401,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/flash","display_name":"Flash (photography)","score":0.7411776185035706},{"id":"https://openalex.org/keywords/flash-memory","display_name":"Flash memory","score":0.6797288656234741},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6512495875358582},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.6061521768569946},{"id":"https://openalex.org/keywords/embedded-system","display_name":"Embedded system","score":0.4034213423728943},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.3941592872142792},{"id":"https://openalex.org/keywords/computer-graphics","display_name":"Computer graphics (images)","score":0.3432142436504364},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.23894274234771729},{"id":"https://openalex.org/keywords/art","display_name":"Art","score":0.100180983543396}],"concepts":[{"id":"https://openalex.org/C2777526259","wikidata":"https://www.wikidata.org/wiki/Q221836","display_name":"Flash (photography)","level":2,"score":0.7411776185035706},{"id":"https://openalex.org/C2776531357","wikidata":"https://www.wikidata.org/wiki/Q174077","display_name":"Flash memory","level":2,"score":0.6797288656234741},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6512495875358582},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.6061521768569946},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.4034213423728943},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3941592872142792},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.3432142436504364},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.23894274234771729},{"id":"https://openalex.org/C142362112","wikidata":"https://www.wikidata.org/wiki/Q735","display_name":"Art","level":0,"score":0.100180983543396},{"id":"https://openalex.org/C153349607","wikidata":"https://www.wikidata.org/wiki/Q36649","display_name":"Visual arts","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hpca61900.2025.00128","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hpca61900.2025.00128","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Symposium on High Performance Computer Architecture (HPCA)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.5400000214576721,"id":"https://metadata.un.org/sdg/7","display_name":"Affordable and clean energy"}],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":80,"referenced_works":["https://openalex.org/W1544971428","https://openalex.org/W1653864426","https://openalex.org/W1978330912","https://openalex.org/W2010202670","https://openalex.org/W2010390358","https://openalex.org/W2014555481","https://openalex.org/W2033828908","https://openalex.org/W2085355957","https://openalex.org/W2125671371","https://openalex.org/W2322680505","https://openalex.org/W2489439822","https://openalex.org/W2588191434","https://openalex.org/W2614421408","https://openalex.org/W2617220603","https://openalex.org/W2804918778","https://openalex.org/W2899332646","https://openalex.org/W2904929935","https://openalex.org/W2971090917","https://openalex.org/W2973214126","https://openalex.org/W3005720814","https://openalex.org/W3015201175","https://openalex.org/W3016243220","https://openalex.org/W3016542674","https://openalex.org/W3016572198","https://openalex.org/W3039374431","https://openalex.org/W3041587974","https://openalex.org/W3100710793","https://openalex.org/W3127556038","https://openalex.org/W3155143326","https://openalex.org/W3158275024","https://openalex.org/W3158860255","https://openalex.org/W3177060288","https://openalex.org/W3192841685","https://openalex.org/W3206688249","https://openalex.org/W3209462773","https://openalex.org/W3210869213","https://openalex.org/W4206254511","https://openalex.org/W4220918199","https://openalex.org/W4221001402","https://openalex.org/W4249654426","https://openalex.org/W4283364281","https://openalex.org/W4292794148","https://openalex.org/W4297097348","https://openalex.org/W4308083744","https://openalex.org/W4312973582","https://openalex.org/W4317382162","https://openalex.org/W4324292875","https://openalex.org/W4360831749","https://openalex.org/W4360831773","https://openalex.org/W4379622897","https://openalex.org/W4385245566","https://openalex.org/W4388757726","https://openalex.org/W4392427708","https://openalex.org/W4393407018","https://openalex.org/W4393407029","https://openalex.org/W4393407316","https://openalex.org/W4394998968","https://openalex.org/W4395073431","https://openalex.org/W4395112660","https://openalex.org/W4401212102","https://openalex.org/W4402671659","https://openalex.org/W4403635905","https://openalex.org/W6732342490","https://openalex.org/W6749289006","https://openalex.org/W6751086749","https://openalex.org/W6791832985","https://openalex.org/W6811129797","https://openalex.org/W6811340617","https://openalex.org/W6847386241","https://openalex.org/W6850625674","https://openalex.org/W6852962002","https://openalex.org/W6856574389","https://openalex.org/W6859967360","https://openalex.org/W6860155063","https://openalex.org/W6861393904","https://openalex.org/W6861995887","https://openalex.org/W6863097700","https://openalex.org/W6865118152","https://openalex.org/W6871104536","https://openalex.org/W6880398243"],"related_works":["https://openalex.org/W2116397085","https://openalex.org/W2535372975","https://openalex.org/W2537636062","https://openalex.org/W2017101954","https://openalex.org/W1594494193","https://openalex.org/W2378293894","https://openalex.org/W2135436866","https://openalex.org/W1994190181","https://openalex.org/W1492907585","https://openalex.org/W2382858638"],"abstract_inverted_index":{"With":[0],"the":[1,11,57,62,65,75,82,89,112,117,166,177,185,192,201,216,219,231,244,257],"widespread":[2],"use":[3,200],"of":[4,35,45,60,189],"large":[5],"language":[6],"models":[7],"(LLMs),":[8],"and":[9,13,64,81,88,133,161,227,248,274,281],"with":[10,106,251,269],"privacy":[12],"cost":[14,226],"concerns":[15],"on":[16,237],"cloud-based":[17],"services,":[18],"vendors":[19],"are":[20],"now":[21],"pushing":[22],"LLM":[23,279],"inference":[24,34],"to":[25,42,73,128,156,169,215,241,254,271],"consumer":[26,153],"devices.":[27],"However,":[28],"current":[29],"attempts":[30],"only":[31],"enable":[32,129],"real-time":[33,258,267],"low-quality":[36],"small-sized":[37],"LLMs.":[38,190],"Large-sized":[39],"LLMs":[40],"have":[41],"load":[43],"most":[44],"their":[46],"weights":[47],"from":[48],"Flash":[49,78,87,109,118,136,140,154,198,221,246],"storage":[50],"for":[51,175,184,207,278],"every":[52],"execution":[53,58],"iteration,":[54],"which":[55,148,209],"dominates":[56],"time":[59],"both":[61,74],"prefill":[63,194,280],"generation":[66,233,282],"phase.":[67],"This":[68],"performance":[69],"bottleneck":[70],"is":[71,149],"attributed":[72],"low":[76,83,162],"internal":[77,119,173,245],"memory":[79],"bandwidth":[80,85,120,174,214],"transmission":[84,178,213],"between":[86],"Neural":[90],"Processing":[91],"Unit":[92],"(NPU).":[93],"To":[94],"tackle":[95],"these":[96],"two":[97,182,186],"challenges,":[98],"we":[99,115,180,196,235],"present":[100],"Lincoln,":[101],"a":[102],"device-architecture":[103],"co-design":[104],"solution":[105],"LPDDR-interfaced,":[107],"Compute-Enabled":[108],"Memory.":[110],"On":[111,165],"device":[113],"level,":[114,168],"boost":[116],"by":[121],"improving":[122],"upon":[123],"existing":[124,202],"array":[125],"shrinking":[126],"methods,":[127],"lower":[130],"read":[131],"latency":[132,259],"more":[134],"parallel":[135],"planes":[137],"within":[138],"each":[139],"die.":[141],"We":[142],"specifically":[143],"leverage":[144,170],"3D":[145],"hybrid":[146],"bonding,":[147],"already":[150],"adopted":[151],"in":[152],"products,":[155],"maintain":[157],"high":[158],"area":[159,228],"efficiency":[160],"density":[163],"loss.":[164],"architecture":[167],"such":[171],"increased":[172],"resolving":[176],"bottleneck,":[179],"propose":[181],"solutions":[183],"distinct":[187],"phases":[188,283],"For":[191,230],"compute-intensive":[193],"phase,":[195,234],"let":[197],"devices":[199],"high-speed":[203],"LPDDR":[204],"interface":[205],"(originally":[206],"DRAM),":[208],"offers":[210],"much":[211],"higher":[212],"NPU":[217],"than":[218],"conventional":[220,285],"interface,":[222],"while":[223],"maintaining":[224],"good":[225],"efficiency.":[229],"memory-intensive":[232],"rely":[236],"hybrid-bonding-based":[238],"near-Flash":[239],"computing":[240],"fully":[242],"utilize":[243],"bandwidth,":[247],"further":[249],"equip":[250],"speculative":[252],"decoding":[253],"eventually":[255],"reach":[256],"goal.":[260],"Our":[261],"evaluation":[262],"shows":[263],"that":[264],"Lincoln":[265],"enables":[266],"inference,":[268],"up":[270],"$13.23":[272],"\\times$":[273,276],"$254.1":[275],"speedups":[277],"over":[284],"SSD-based":[286],"systems.":[287]},"counts_by_year":[{"year":2026,"cited_by_count":2},{"year":2025,"cited_by_count":3}],"updated_date":"2026-03-13T16:22:10.518609","created_date":"2025-10-10T00:00:00"}
