{"id":"https://openalex.org/W4416429485","doi":"https://doi.org/10.1109/iccad66269.2025.11240867","title":"AccelStack: A Cost-Driven Analysis of 3D-Stacked LLM Accelerators","display_name":"AccelStack: A Cost-Driven Analysis of 3D-Stacked LLM Accelerators","publication_year":2025,"publication_date":"2025-10-26","ids":{"openalex":"https://openalex.org/W4416429485","doi":"https://doi.org/10.1109/iccad66269.2025.11240867"},"language":"en","primary_location":{"id":"doi:10.1109/iccad66269.2025.11240867","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccad66269.2025.11240867","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Conference On Computer Aided Design (ICCAD)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5100662676","display_name":"Chen Bai","orcid":"https://orcid.org/0000-0002-1742-0090"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Chen Bai","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000911398","display_name":"Xin Fan","orcid":"https://orcid.org/0009-0008-4658-8192"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Xin Fan","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103138440","display_name":"Zhenhua Zhu","orcid":"https://orcid.org/0009-0007-9259-7180"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Zhenhua Zhu","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100441765","display_name":"Wei Zhang","orcid":"https://orcid.org/0000-0003-2255-5405"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Wei Zhang","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5029566713","display_name":"Yuan Xie","orcid":"https://orcid.org/0000-0001-5399-0501"},"institutions":[{"id":"https://openalex.org/I200769079","display_name":"Hong Kong University of Science and Technology","ror":"https://ror.org/00q4vv597","country_code":"HK","type":"education","lineage":["https://openalex.org/I200769079"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Yuan Xie","raw_affiliation_strings":["The Hong Kong University of Science and Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"The Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I200769079"],"apc_list":null,"apc_paid":null,"fwci":6.9408,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.97333797,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":97,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"9"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11527","display_name":"3D IC and TSV technologies","score":0.8695999979972839,"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"}},"topics":[{"id":"https://openalex.org/T11527","display_name":"3D IC and TSV technologies","score":0.8695999979972839,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.05139999836683273,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.025800000876188278,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/leverage","display_name":"Leverage (statistics)","score":0.5397999882698059},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.40639999508857727},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.3849000036716461},{"id":"https://openalex.org/keywords/architecture","display_name":"Architecture","score":0.365200012922287},{"id":"https://openalex.org/keywords/memory-bandwidth","display_name":"Memory bandwidth","score":0.28929999470710754},{"id":"https://openalex.org/keywords/power","display_name":"Power (physics)","score":0.2770000100135803}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6509000062942505},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.5397999882698059},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.45590001344680786},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.4260999858379364},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.40639999508857727},{"id":"https://openalex.org/C201995342","wikidata":"https://www.wikidata.org/wiki/Q682496","display_name":"Systems engineering","level":1,"score":0.38769999146461487},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.3849000036716461},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.365200012922287},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3596000075340271},{"id":"https://openalex.org/C200601418","wikidata":"https://www.wikidata.org/wiki/Q2193887","display_name":"Reliability engineering","level":1,"score":0.3531000018119812},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3192000091075897},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.30140000581741333},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2939999997615814},{"id":"https://openalex.org/C188045654","wikidata":"https://www.wikidata.org/wiki/Q17148339","display_name":"Memory bandwidth","level":2,"score":0.28929999470710754},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2793999910354614},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.2770000100135803},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.2741999924182892},{"id":"https://openalex.org/C12186640","wikidata":"https://www.wikidata.org/wiki/Q6815743","display_name":"Memory model","level":3,"score":0.26440000534057617},{"id":"https://openalex.org/C2982832238","wikidata":"https://www.wikidata.org/wiki/Q5531640","display_name":"General purpose","level":2,"score":0.2563000023365021},{"id":"https://openalex.org/C2779602883","wikidata":"https://www.wikidata.org/wiki/Q15544750","display_name":"Memory architecture","level":2,"score":0.2556999921798706}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/iccad66269.2025.11240867","is_oa":false,"landing_page_url":"https://doi.org/10.1109/iccad66269.2025.11240867","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE/ACM International Conference On Computer Aided Design (ICCAD)","raw_type":"proceedings-article"},{"id":"pmh:oai:repository.hkust.edu.hk:1783.1-171262","is_oa":false,"landing_page_url":"http://repository.hkust.edu.hk/ir/Record/1783.1-171262","pdf_url":null,"source":{"id":"https://openalex.org/S4306401796","display_name":"Rare & Special e-Zone (The Hong Kong University of Science and Technology)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I200769079","host_organization_name":"Hong Kong University of Science and Technology","host_organization_lineage":["https://openalex.org/I200769079"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Conference paper"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":29,"referenced_works":["https://openalex.org/W2042120213","https://openalex.org/W2162009815","https://openalex.org/W2518432791","https://openalex.org/W2532719504","https://openalex.org/W2970073801","https://openalex.org/W3130554079","https://openalex.org/W3157657667","https://openalex.org/W3166510811","https://openalex.org/W3188178661","https://openalex.org/W3192636176","https://openalex.org/W4206364646","https://openalex.org/W4220769636","https://openalex.org/W4220972538","https://openalex.org/W4283076324","https://openalex.org/W4285047926","https://openalex.org/W4293023438","https://openalex.org/W4324292875","https://openalex.org/W4376130831","https://openalex.org/W4380874786","https://openalex.org/W4385192563","https://openalex.org/W4388031315","https://openalex.org/W4388979610","https://openalex.org/W4394693961","https://openalex.org/W4401211711","https://openalex.org/W4407692602","https://openalex.org/W4407695247","https://openalex.org/W4411689103","https://openalex.org/W4414197293","https://openalex.org/W4414281281"],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"show":[4,82],"viability":[5],"for":[6,41,52,71],"artificial":[7],"general":[8],"intelligence":[9],"(AGI)":[10],"with":[11,106],"high":[12],"computing":[13],"power":[14],"and":[15,27,76,89,97],"memory":[16,25],"bandwidth":[17],"demands.":[18],"While":[19],"existing":[20],"LLM":[21,43,104],"accelerators":[22,84],"leverage":[23],"high-bandwidth":[24],"(HBM)":[26],"2.5D":[28],"packaging":[29],"to":[30,87],"address":[31],"the":[32,53],"challenge,":[33],"emerging":[34],"hybrid":[35],"bonding":[36],"techniques":[37],"unlock":[38],"new":[39,54],"opportunities":[40],"3D-stacked":[42,83],"accelerators.":[44],"This":[45],"paper":[46],"proposes":[47],"AccelStack,":[48],"a":[49,60,68],"cost-driven":[50],"analysis":[51],"architecture":[55],"via":[56],"two":[57],"innovations.":[58],"First,":[59],"performance":[61],"model":[62,70],"capturing":[63],"memory-on-logic":[64],"is":[65,79],"presented.":[66],"Second,":[67],"cost":[69],"die-on-die":[72],"(DoD),":[73],"die-on-wafer":[74],"(DoW),":[75],"wafer-on-wafer":[77],"(WoW)":[78],"proposed.":[80],"Evaluations":[81],"achieve":[85],"up":[86],"7.17\u00d7":[88],"2.09\u00d7":[90],"faster":[91],"inference":[92],"than":[93],"NVIDIA":[94],"A100":[95],"(FP16)":[96],"H100":[98],"(FP8)":[99],"simulation":[100],"results":[101],"across":[102],"various":[103],"workloads,":[105],"chiplet-based":[107],"designs":[108],"reducing":[109],"recurring":[110],"engineering":[111],"costs":[112],"by":[113],"38.09%":[114],"versus":[115],"monolithic":[116],"implementations.":[117]},"counts_by_year":[{"year":2026,"cited_by_count":2}],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2025-11-20T00:00:00"}