{"id":"https://openalex.org/W4413145360","doi":"https://doi.org/10.1109/ipdpsw66978.2025.00151","title":"Evaluating Expansion Memory for Optimizer State Offloading for Large Transformer Models","display_name":"Evaluating Expansion Memory for Optimizer State Offloading for Large Transformer Models","publication_year":2025,"publication_date":"2025-06-03","ids":{"openalex":"https://openalex.org/W4413145360","doi":"https://doi.org/10.1109/ipdpsw66978.2025.00151"},"language":"en","primary_location":{"id":"doi:10.1109/ipdpsw66978.2025.00151","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdpsw66978.2025.00151","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://hal.science/hal-04996269","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5017707472","display_name":"Moiz Arif","orcid":"https://orcid.org/0000-0002-8823-5562"},"institutions":[{"id":"https://openalex.org/I11912373","display_name":"Micron (United States)","ror":"https://ror.org/02fv52296","country_code":"US","type":"company","lineage":["https://openalex.org/I11912373"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Moiz Arif","raw_affiliation_strings":["Micron Technology Inc,Austin,TX,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology Inc,Austin,TX,USA","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5049218913","display_name":"Avinash Maurya","orcid":"https://orcid.org/0000-0002-8200-0148"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Avinash Maurya","raw_affiliation_strings":["Argonne National Laboratory,Lemont,IL,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Argonne National Laboratory,Lemont,IL,USA","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000418371","display_name":"Sudharshan S. Vazhkudai","orcid":"https://orcid.org/0000-0002-8596-5033"},"institutions":[{"id":"https://openalex.org/I11912373","display_name":"Micron (United States)","ror":"https://ror.org/02fv52296","country_code":"US","type":"company","lineage":["https://openalex.org/I11912373"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sudharshan Vazhkudai","raw_affiliation_strings":["Micron Technology Inc,Austin,TX,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology Inc,Austin,TX,USA","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5085745891","display_name":"Bogdan Nicolae","orcid":"https://orcid.org/0000-0002-0661-7509"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bogdan Nicolae","raw_affiliation_strings":["Argonne National Laboratory,Lemont,IL,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Argonne National Laboratory,Lemont,IL,USA","institution_ids":["https://openalex.org/I1282105669"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":5.5066,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.95396154,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"970","last_page":"977"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11343","display_name":"Power Transformer Diagnostics and Insulation","score":0.9613999724388123,"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/T11343","display_name":"Power Transformer Diagnostics and Insulation","score":0.9613999724388123,"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/T11222","display_name":"Magnetic Properties and Applications","score":0.9581000208854675,"subfield":{"id":"https://openalex.org/subfields/2504","display_name":"Electronic, Optical and Magnetic Materials"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10472","display_name":"Semiconductor materials and devices","score":0.932699978351593,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.7091845870018005},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.608309805393219},{"id":"https://openalex.org/keywords/electrical-engineering","display_name":"Electrical engineering","score":0.18351513147354126},{"id":"https://openalex.org/keywords/voltage","display_name":"Voltage","score":0.12310194969177246},{"id":"https://openalex.org/keywords/engineering","display_name":"Engineering","score":0.10818731784820557}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7091845870018005},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.608309805393219},{"id":"https://openalex.org/C119599485","wikidata":"https://www.wikidata.org/wiki/Q43035","display_name":"Electrical engineering","level":1,"score":0.18351513147354126},{"id":"https://openalex.org/C165801399","wikidata":"https://www.wikidata.org/wiki/Q25428","display_name":"Voltage","level":2,"score":0.12310194969177246},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.10818731784820557}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/ipdpsw66978.2025.00151","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdpsw66978.2025.00151","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW)","raw_type":"proceedings-article"},{"id":"pmh:oai:HAL:hal-04996269v1","is_oa":true,"landing_page_url":"https://hal.science/hal-04996269","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"HPAI4S'25: HPC for AI Foundation Models & LLMs for Science (co-located with IPDPS'25), Jun 2025, Milan, Italy","raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"pmh:oai:HAL:hal-04996269v1","is_oa":true,"landing_page_url":"https://hal.science/hal-04996269","pdf_url":null,"source":{"id":"https://openalex.org/S4306402512","display_name":"HAL (Le Centre pour la Communication Scientifique Directe)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I1294671590","host_organization_name":"Centre National de la Recherche Scientifique","host_organization_lineage":["https://openalex.org/I1294671590"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"HPAI4S'25: HPC for AI Foundation Models & LLMs for Science (co-located with IPDPS'25), Jun 2025, Milan, Italy","raw_type":"info:eu-repo/semantics/conferenceObject"},"sustainable_development_goals":[{"display_name":"Affordable and clean energy","id":"https://metadata.un.org/sdg/7","score":0.7599999904632568}],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320306084","display_name":"U.S. Department of Energy","ror":"https://ror.org/01bj3aw27"},{"id":"https://openalex.org/F4320332359","display_name":"Office of Science","ror":"https://ror.org/00mmn6b08"},{"id":"https://openalex.org/F4320337377","display_name":"Office of Advanced Cyberinfrastructure","ror":"https://ror.org/04nh1dc89"},{"id":"https://openalex.org/F4320337506","display_name":"Advanced Scientific Computing Research","ror":"https://ror.org/0012c7r22"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2899084033","https://openalex.org/W2748952813","https://openalex.org/W2390279801","https://openalex.org/W4391913857","https://openalex.org/W2358668433","https://openalex.org/W4396701345","https://openalex.org/W2376932109","https://openalex.org/W2001405890","https://openalex.org/W4396696052"],"abstract_inverted_index":{"Transformers":[0],"and":[1,23,29,49,85,114,126,180],"large":[2,107],"language":[3,27],"models":[4,41,156],"(LLMs)":[5],"have":[6,24,75],"seen":[7],"rapid":[8],"adoption":[9],"across":[10],"various":[11],"domains,":[12],"with":[13],"their":[14],"sizes":[15],"exploding":[16],"to":[17,82,142,145],"hundreds":[18],"of":[19,21,39,59,68,95,116,128,153],"billions":[20],"parameters":[22],"revolutionized":[25],"natural":[26],"processing":[28],"other":[30],"artificial":[31],"intelligence":[32],"(AI)":[33],"applications.":[34],"However,":[35],"the":[36,46,57,65,93,112,123,151,175],"increasing":[37],"complexity":[38],"such":[40],"poses":[42],"significant":[43],"challenges":[44],"for":[45,105,170,177],"available":[47],"high-bandwidth,":[48],"main":[50,83],"memory":[51,66,84,103,168],"resources.":[52],"One":[53],"critical":[54],"issue":[55],"is":[56],"growth":[58],"optimizer":[60,80,108],"states,":[61],"which":[62],"often":[63],"exceed":[64],"capacity":[67],"GPUs.":[69],"To":[70],"address":[71],"this,":[72],"several":[73],"strategies":[74],"been":[76],"explored,":[77],"including":[78],"offloading":[79,106,136],"states":[81],"NVMe":[86,120,146],"storage.":[87],"In":[88],"this":[89],"paper,":[90],"we":[91],"investigate":[92],"use":[94],"compute":[96],"express":[97],"link":[98],"(CXL)":[99],"as":[100],"an":[101],"expansion":[102],"solution":[104],"states.":[109],"We":[110],"compare":[111],"performance":[113],"efficiency":[115],"CXL":[117],"against":[118],"next-generation":[119],"devices,":[121],"highlighting":[122],"potential":[124],"benefits":[125],"limitations":[127],"each":[129],"approach.":[130],"Our":[131],"findings":[132],"show":[133],"that":[134],"CXL-based":[135],"reduces":[137],"training":[138,152],"time":[139],"by":[140],"up":[141],"40%":[143],"compared":[144],"offloading,":[147],"while":[148],"also":[149],"enabling":[150],"significantly":[154],"larger":[155],"than":[157],"DRAM":[158],"only":[159],"offloading.":[160],"This":[161],"research":[162],"provides":[163],"valuable":[164],"insights":[165],"into":[166],"optimizing":[167],"management":[169],"large-scale":[171],"transformer":[172],"models,":[173],"paving":[174],"way":[176],"more":[178],"efficient":[179],"scalable":[181],"AI":[182],"systems.":[183]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-08-26T07:47:46.906454","created_date":"2025-10-10T00:00:00"}