{"id":"https://openalex.org/W4414198888","doi":"https://doi.org/10.1109/dac63849.2025.11133230","title":"AttenPIM: Accelerating LLM Attention with Dual-mode GEMV in Processing-in-Memory","display_name":"AttenPIM: Accelerating LLM Attention with Dual-mode GEMV in Processing-in-Memory","publication_year":2025,"publication_date":"2025-06-22","ids":{"openalex":"https://openalex.org/W4414198888","doi":"https://doi.org/10.1109/dac63849.2025.11133230"},"language":"en","primary_location":{"id":"doi:10.1109/dac63849.2025.11133230","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11133230","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","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/A5073749486","display_name":"Liyan Chen","orcid":"https://orcid.org/0000-0001-9693-5377"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":true,"raw_author_name":"Liyan Chen","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5066692868","display_name":"Dongxu Lyu","orcid":"https://orcid.org/0000-0001-6826-2670"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Dongxu Lyu","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100330782","display_name":"Zhenyu Li","orcid":"https://orcid.org/0000-0003-2112-9834"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhenyu Li","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008057183","display_name":"Jianfei Jiang","orcid":"https://orcid.org/0000-0002-5521-6197"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jianfei Jiang","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063554298","display_name":"Qin Wang","orcid":"https://orcid.org/0000-0002-6559-5207"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Qin Wang","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103236320","display_name":"Zhigang Mao","orcid":"https://orcid.org/0000-0001-9431-9853"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhigang Mao","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5045693138","display_name":"Naifeng Jing","orcid":"https://orcid.org/0000-0001-8417-5796"},"institutions":[{"id":"https://openalex.org/I183067930","display_name":"Shanghai Jiao Tong University","ror":"https://ror.org/0220qvk04","country_code":"CN","type":"education","lineage":["https://openalex.org/I183067930"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Naifeng Jing","raw_affiliation_strings":["Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China"],"affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University,Department of Micro/Nano Electronics,Shanghai,China","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5073749486"],"corresponding_institution_ids":["https://openalex.org/I183067930"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.12631476,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.9505000114440918,"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"}},"topics":[{"id":"https://openalex.org/T12676","display_name":"Machine Learning and ELM","score":0.9505000114440918,"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/T10502","display_name":"Advanced Memory and Neural Computing","score":0.942300021648407,"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/T10320","display_name":"Neural Networks and Applications","score":0.9307000041007996,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/concurrency","display_name":"Concurrency","score":0.6690000295639038},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6409000158309937},{"id":"https://openalex.org/keywords/computation-offloading","display_name":"Computation offloading","score":0.6157000064849854},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.5699999928474426},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.567799985408783},{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.5648000240325928},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.4731000065803528},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.43369999527931213}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.847599983215332},{"id":"https://openalex.org/C193702766","wikidata":"https://www.wikidata.org/wiki/Q1414548","display_name":"Concurrency","level":2,"score":0.6690000295639038},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6409000158309937},{"id":"https://openalex.org/C2781041963","wikidata":"https://www.wikidata.org/wiki/Q18348618","display_name":"Computation offloading","level":4,"score":0.6157000064849854},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.5699999928474426},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.567799985408783},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.5648000240325928},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.5109999775886536},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.4731000065803528},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.43369999527931213},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.42489999532699585},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.40639999508857727},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.39469999074935913},{"id":"https://openalex.org/C2780165032","wikidata":"https://www.wikidata.org/wiki/Q16869822","display_name":"Energy consumption","level":2,"score":0.3528999984264374},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.33149999380111694},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.31119999289512634},{"id":"https://openalex.org/C150495011","wikidata":"https://www.wikidata.org/wiki/Q128392","display_name":"Concurrent computing","level":2,"score":0.3043999969959259},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.30399999022483826},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.29910001158714294},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.29030001163482666},{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.2773999869823456},{"id":"https://openalex.org/C53833338","wikidata":"https://www.wikidata.org/wiki/Q1061424","display_name":"Context switch","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C2777735758","wikidata":"https://www.wikidata.org/wiki/Q817765","display_name":"Path (computing)","level":2,"score":0.26429998874664307},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2565999925136566},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.2554999887943268}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/dac63849.2025.11133230","is_oa":false,"landing_page_url":"https://doi.org/10.1109/dac63849.2025.11133230","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 62nd ACM/IEEE Design Automation Conference (DAC)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":15,"referenced_works":["https://openalex.org/W2982008795","https://openalex.org/W3006586535","https://openalex.org/W3100710793","https://openalex.org/W3210580311","https://openalex.org/W4249322926","https://openalex.org/W4297097348","https://openalex.org/W4385834084","https://openalex.org/W4387064057","https://openalex.org/W4392427708","https://openalex.org/W4393407040","https://openalex.org/W4394998968","https://openalex.org/W4395106409","https://openalex.org/W4401880111","https://openalex.org/W4402475900","https://openalex.org/W4402671766"],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"have":[4],"demonstrated":[5],"unprecedented":[6],"generative":[7],"performance":[8],"across":[9],"a":[10,102,136],"wide":[11],"range":[12],"of":[13,37],"applications.":[14],"While":[15],"recent":[16],"heterogeneous":[17],"architectures":[18],"attempt":[19],"to":[20,152,170,194],"address":[21,96],"the":[22],"memory-bound":[23],"bottleneck":[24],"from":[25,43,84],"attention":[26,38,108],"computations":[27,123],"by":[28,135,189],"processing-in-memory":[29],"(PIM)":[30],"offloading,":[31],"they":[32],"overlook":[33],"two":[34,195],"critical":[35],"characteristics":[36],"GEMVs":[39],"that":[40,51,178],"distinguish":[41],"them":[42],"traditional":[44],"PIM":[45,74,140,197],"scenarios:":[46],"(1)":[47],"dynamic":[48,162],"matrix":[49],"dimensions":[50],"scale":[52],"with":[53,124],"token":[54],"length,":[55],"and":[56,67,121,128,149,156,164,185],"(2)":[57],"distinct":[58,92],"GEMV":[59,93],"patterns":[60],"between":[61],"score":[62,120],"computation":[63,69],"($Q":[64],"\\times":[65,71],"K_{t}$)":[66],"context":[68,122],"($S":[70],"V$).":[72],"Existing":[73],"designs,":[75],"employing":[76],"either":[77],"uniform":[78],"or":[79,91],"transposed":[80],"computing":[81,116],"modes,":[82],"suffer":[83],"inefficiencies":[85],"in":[86],"newly":[87],"generated":[88],"element":[89],"preparation":[90],"execution.":[94],"To":[95],"these":[97],"limitations,":[98],"we":[99,113,146],"propose":[100,114],"AttenPIM,":[101],"software-hardware":[103],"co-design":[104],"for":[105,119,131],"efficient":[106],"PIM-based":[107],"acceleration.":[109],"For":[110,143],"bank-level":[111],"execution,":[112,145],"dual-mode":[115],"modes":[117],"tailored":[118],"PIM-oriented":[125],"data":[126],"layouts":[127],"execution":[129],"flows":[130],"KV":[132],"storage,":[133],"supported":[134],"low-cost":[137],"configurable":[138],"per-bank":[139],"unit":[141],"(PU).":[142],"system-level":[144],"leverage":[147],"token-level":[148],"head-level":[150],"concurrency":[151],"ensure":[153],"workload":[154],"balance":[155],"maximize":[157],"bank":[158],"PU":[159],"parallelism.":[160],"Furthermore,":[161],"allocation":[163],"kernel":[165],"fusion":[166],"methods":[167],"are":[168],"proposed":[169],"further":[171],"minimize":[172],"memory":[173],"overhead.":[174],"Experimental":[175],"results":[176],"demonstrate":[177],"AttenPIM":[179],"achieves":[180],"$1.13":[181],"\\times-5.26":[182],"\\times$":[183],"speedup":[184],"reduces":[186],"energy":[187],"consumption":[188],"17":[190],"%-49":[191],"%":[192],"compared":[193],"state-of-the-art":[196],"baselines.":[198]},"counts_by_year":[],"updated_date":"2026-04-09T08:11:56.329763","created_date":"2025-10-10T00:00:00"}