{"id":"https://openalex.org/W4407466309","doi":"https://doi.org/10.1109/lca.2025.3532682","title":"Cost-Effective Extension of DRAM-PIM for Group-Wise LLM Quantization","display_name":"Cost-Effective Extension of DRAM-PIM for Group-Wise LLM Quantization","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4407466309","doi":"https://doi.org/10.1109/lca.2025.3532682"},"language":"en","primary_location":{"id":"doi:10.1109/lca.2025.3532682","is_oa":true,"landing_page_url":"https://doi.org/10.1109/lca.2025.3532682","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1109/lca.2025.3532682","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5062044183","display_name":"B. J. Kim","orcid":null},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":true,"raw_author_name":"Byeori Kim","raw_affiliation_strings":["Department of Computer Science and Engineering, POSTECH, Pohang, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, POSTECH, Pohang, South Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5116568538","display_name":"Changhun Lee","orcid":"https://orcid.org/0000-0002-6043-034X"},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Changhun Lee","raw_affiliation_strings":["Department of Convergence IT Engineering, POSTECH, Pohang, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Convergence IT Engineering, POSTECH, Pohang, South Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100334897","display_name":"Gwangsun Kim","orcid":"https://orcid.org/0000-0001-5749-5794"},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Gwangsun Kim","raw_affiliation_strings":["Department of Computer Science and Engineering, POSTECH, Pohang, South Korea"],"affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, POSTECH, Pohang, South Korea","institution_ids":["https://openalex.org/I123900574"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5064884832","display_name":"Eunhyeok Park","orcid":"https://orcid.org/0000-0002-7331-9819"},"institutions":[{"id":"https://openalex.org/I123900574","display_name":"Pohang University of Science and Technology","ror":"https://ror.org/04xysgw12","country_code":"KR","type":"education","lineage":["https://openalex.org/I123900574"]}],"countries":["KR"],"is_corresponding":false,"raw_author_name":"Eunhyeok Park","raw_affiliation_strings":["Graduate School of Artificial Intelligence, POSTECH, Pohang, South Korea"],"affiliations":[{"raw_affiliation_string":"Graduate School of Artificial Intelligence, POSTECH, Pohang, South Korea","institution_ids":["https://openalex.org/I123900574"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":4,"corresponding_author_ids":["https://openalex.org/A5062044183"],"corresponding_institution_ids":["https://openalex.org/I123900574"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.02367387,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"24","issue":"1","first_page":"53","last_page":"56"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11206","display_name":"Model Reduction and Neural Networks","score":0.9829999804496765,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11206","display_name":"Model Reduction and Neural Networks","score":0.9829999804496765,"subfield":{"id":"https://openalex.org/subfields/3109","display_name":"Statistical and Nonlinear Physics"},"field":{"id":"https://openalex.org/fields/31","display_name":"Physics and Astronomy"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T11263","display_name":"Electromagnetic Simulation and Numerical Methods","score":0.9660000205039978,"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"}},{"id":"https://openalex.org/T11034","display_name":"Digital Filter Design and Implementation","score":0.9607999920845032,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/dram","display_name":"Dram","score":0.799813985824585},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.67087322473526},{"id":"https://openalex.org/keywords/extension","display_name":"Extension (predicate logic)","score":0.6613321304321289},{"id":"https://openalex.org/keywords/quantization","display_name":"Quantization (signal processing)","score":0.5344125032424927},{"id":"https://openalex.org/keywords/group","display_name":"Group (periodic table)","score":0.4387732148170471},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3692406713962555},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3327930271625519},{"id":"https://openalex.org/keywords/programming-language","display_name":"Programming language","score":0.21742430329322815},{"id":"https://openalex.org/keywords/computer-hardware","display_name":"Computer hardware","score":0.20130997896194458},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.14319166541099548}],"concepts":[{"id":"https://openalex.org/C7366592","wikidata":"https://www.wikidata.org/wiki/Q1255620","display_name":"Dram","level":2,"score":0.799813985824585},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.67087322473526},{"id":"https://openalex.org/C2778029271","wikidata":"https://www.wikidata.org/wiki/Q5421931","display_name":"Extension (predicate logic)","level":2,"score":0.6613321304321289},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.5344125032424927},{"id":"https://openalex.org/C2781311116","wikidata":"https://www.wikidata.org/wiki/Q83306","display_name":"Group (periodic table)","level":2,"score":0.4387732148170471},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3692406713962555},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3327930271625519},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.21742430329322815},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.20130997896194458},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.14319166541099548},{"id":"https://openalex.org/C178790620","wikidata":"https://www.wikidata.org/wiki/Q11351","display_name":"Organic chemistry","level":1,"score":0.0},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/lca.2025.3532682","is_oa":true,"landing_page_url":"https://doi.org/10.1109/lca.2025.3532682","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1109/lca.2025.3532682","is_oa":true,"landing_page_url":"https://doi.org/10.1109/lca.2025.3532682","pdf_url":null,"source":{"id":"https://openalex.org/S17643076","display_name":"IEEE Computer Architecture Letters","issn_l":"1556-6056","issn":["1556-6056","1556-6064","2473-2575"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Computer Architecture Letters","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320315121","display_name":"Samsung Advanced Institute of Technology","ror":null}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":9,"referenced_works":["https://openalex.org/W2162639668","https://openalex.org/W2896090304","https://openalex.org/W3119993810","https://openalex.org/W3189166979","https://openalex.org/W4280496502","https://openalex.org/W4385573671","https://openalex.org/W4392427708","https://openalex.org/W4394998968","https://openalex.org/W6877022392"],"related_works":["https://openalex.org/W3120961607","https://openalex.org/W4401568740","https://openalex.org/W3148568549","https://openalex.org/W2098207691","https://openalex.org/W1648516568","https://openalex.org/W361036515","https://openalex.org/W2161286015","https://openalex.org/W2269474412","https://openalex.org/W2433923775","https://openalex.org/W80860381"],"abstract_inverted_index":{"Processing-in-Memory":[0],"(PIM)":[1],"is":[2,100],"emerging":[3],"as":[4],"a":[5,168,180,191],"promising":[6],"next-generation":[7],"hardware":[8],"to":[9,42,62,105,131,145,161,186],"address":[10],"memory":[11,22,109],"bottlenecks":[12],"in":[13,48,176],"large":[14,31],"language":[15],"model":[16,58],"(LLM)":[17],"inference":[18],"by":[19],"leveraging":[20],"internal":[21],"bandwidth,":[23],"enabling":[24,60],"more":[25],"energy-efficient":[26],"on-device":[27],"AI.":[28],"However,":[29],"LLMs\u2019":[30],"footprint":[32],"poses":[33],"significant":[34],"challenges":[35],"for":[36,88,172],"accelerating":[37],"them":[38],"on":[39,135],"PIM":[40,77,136,164],"due":[41,104],"limited":[43],"available":[44],"space.":[45],"Recent":[46],"advances":[47],"weight-only":[49],"quantization,":[50],"especially":[51],"group-wise":[52],"weight":[53],"quantization":[54],"(GWQ),":[55],"reduce":[56],"LLM":[57],"sizes,":[59],"parameters":[61],"be":[63],"stored":[64],"at":[65],"4-bit":[66,177],"precision":[67],"or":[68],"lower":[69],"with":[70,113,137,179],"minimal":[71,114,138],"accuracy":[72],"loss.":[73],"Despite":[74],"this,":[75],"current":[76],"architectures":[78],"experience":[79],"performance":[80,193],"degradation":[81],"when":[82],"handling":[83],"the":[84,106,152,162],"additional":[85],"computations":[86],"required":[87],"quantized":[89],"weights.":[90],"While":[91],"incorporating":[92],"extra":[93],"logic":[94],"could":[95],"mitigate":[96],"this":[97],"degradation,":[98],"it":[99],"often":[101],"prohibitively":[102],"expensive":[103],"constraints":[107],"of":[108,183],"technology,":[110],"necessitating":[111],"solutions":[112],"area":[115,142,159],"overhead.":[116],"This":[117],"work":[118],"introduces":[119],"two":[120],"key":[121],"innovations:":[122],"1)":[123],"scale":[124],"cascading,":[125],"and":[126,141,166,174],"2)":[127],"an":[128],"INT2FP":[129],"converter,":[130],"support":[132],"GWQ-applied":[133],"LLMs":[134],"dequantization":[139,173],"latency":[140,170],"overhead":[143,160,171],"compared":[144,185],"FP16":[146,187],"GEMV.":[147],"Experimental":[148],"results":[149],"show":[150],"that":[151],"proposed":[153],"approach":[154],"adds":[155],"less":[156],"than":[157],"0.6%":[158],"existing":[163],"unit":[165],"achieves":[167],"7%":[169],"GEMV":[175],"GWQ":[178],"group":[181],"size":[182],"128,":[184],"GEMV,":[188],"while":[189],"offering":[190],"1.55\u00d7":[192],"gain":[194],"over":[195],"baseline":[196],"dequantization.":[197]},"counts_by_year":[],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}