{"id":"https://openalex.org/W7148522296","doi":"https://doi.org/10.1109/hipcw66559.2025.00021","title":"Evaluating CXL Memory Pooling for Scalable LLM Inference","display_name":"Evaluating CXL Memory Pooling for Scalable LLM Inference","publication_year":2025,"publication_date":"2025-12-17","ids":{"openalex":"https://openalex.org/W7148522296","doi":"https://doi.org/10.1109/hipcw66559.2025.00021"},"language":null,"primary_location":{"id":"doi:10.1109/hipcw66559.2025.00021","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hipcw66559.2025.00021","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 32nd International Conference on High Performance Computing, Data and Analytics Workshop (HiPCW)","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/A5132746152","display_name":"Sai Krishna Vemuri","orcid":null},"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":"Sai Krishna Vemuri","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132758996","display_name":"Venkata Ravi Shankar Jonnalagadda","orcid":null},"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":"Venkata Ravi Shankar Jonnalagadda","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132752438","display_name":"Ajay Joshi","orcid":null},"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":"Ajay Joshi","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5046552998","display_name":"Rohit Sindhu","orcid":null},"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":"Rohit Sindhu","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132779360","display_name":"Vijay Kumar Motagi","orcid":null},"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":"Vijay Kumar Motagi","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132777971","display_name":"Amandeep Singh","orcid":null},"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":"Amandeep Singh","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5132771243","display_name":"Praveen Kumar Rukmangada","orcid":null},"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":"Praveen Kumar Rukmangada","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5028650344","display_name":"Sachin Shaw","orcid":"https://orcid.org/0000-0002-5456-8596"},"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":"Sachin Shaw","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5132780338","display_name":"Vishal Tanna","orcid":null},"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":"Vishal Tanna","raw_affiliation_strings":["Micron Technology"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Micron Technology","institution_ids":["https://openalex.org/I11912373"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I11912373"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.7512134,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"75","last_page":"79"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.19099999964237213,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.19099999964237213,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.08940000087022781,"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/T12326","display_name":"Network Packet Processing and Optimization","score":0.050599999725818634,"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/pooling","display_name":"Pooling","score":0.5849000215530396},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5807999968528748},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.4871000051498413},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.2840999960899353},{"id":"https://openalex.org/keywords/pattern-recognition","display_name":"Pattern recognition (psychology)","score":0.25519999861717224}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7081000208854675},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6049000024795532},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.5849000215530396},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5807999968528748},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.4871000051498413},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.47200000286102295},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.33660000562667847},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2840999960899353},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.25519999861717224},{"id":"https://openalex.org/C160234255","wikidata":"https://www.wikidata.org/wiki/Q812535","display_name":"Bayesian inference","level":3,"score":0.25429999828338623}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/hipcw66559.2025.00021","is_oa":false,"landing_page_url":"https://doi.org/10.1109/hipcw66559.2025.00021","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 32nd International Conference on High Performance Computing, Data and Analytics Workshop (HiPCW)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":4,"referenced_works":["https://openalex.org/W4318767786","https://openalex.org/W4387321091","https://openalex.org/W4405348560","https://openalex.org/W4408738433"],"related_works":[],"abstract_inverted_index":{"Large-context":[0],"LLM":[1,95],"inference":[2,100],"increasingly":[3],"faces":[4],"bottlenecks":[5],"in":[6,82,202],"Key-Value":[7],"(KV)":[8],"cache":[9,37,104],"capacity":[10,24],"and":[11,69,79,89,102,110,116,137,152,196],"bandwidth":[12],"rather":[13],"than":[14],"raw":[15],"compute.":[16],"While":[17],"on-package":[18],"HBM":[19,35,133,138],"delivers":[20],"exceptional":[21],"bandwidth,":[22],"its":[23],"is":[25],"limited;":[26],"when":[27,88],"long":[28],"prompts":[29],"or":[30,43,48],"multi-tenant":[31],"workloads":[32],"exceed":[33],"the":[34,99,166,185],"KV":[36,103],"budget,":[38],"operators":[39],"turn":[40],"to":[41,45,180],"re-computation":[42],"offloading":[44],"system":[46],"memory":[47,63,92],"storage\u2014both":[49],"with":[50,131,192],"significant":[51],"performance":[52],"costs.":[53],"Compute":[54],"Express":[55],"Link":[56],"(CXL)":[57],"introduces":[58],"a":[59,118],"flexible,":[60],"low-latency,":[61],"load/store":[62],"tier":[64],"that":[65,128,158],"can":[66],"be":[67],"pooled":[68,141],"shared":[70],"across":[71],"nodes":[72],"using":[73],"PCIe":[74],"6.0":[75],"PHY,":[76],"multi-level":[77],"switching,":[78],"fabric":[80],"features":[81],"CXL":[83,91,114,140,153,159,200],"3.x.":[84],"This":[85],"paper":[86],"examines":[87],"why":[90],"pooling":[93,160],"helps":[94],"serving:":[96],"we":[97],"explain":[98],"pipeline":[101],"growth":[105],"dynamics,":[106],"discuss":[107],"architectural":[108],"advantages":[109],"deployment":[111],"considerations":[112],"for":[113,198],"pooling,":[115],"present":[117],"compact":[119],"analytical":[120],"model":[121,156],"of":[122,188],"per-accelerator":[123],"throughput":[124,164],"vs.":[125],"context":[126],"length":[127],"compares":[129],"HBM-only":[130],"re-computation,":[132],"+":[134,139],"distributed":[135,181],"DDR,":[136,182],"memory.":[142],"With":[143],"realistic":[144],"parameters":[145],"from":[146],"contemporary":[147],"GPUs,":[148],"800":[149],"GbE":[150],"networks,":[151],"fabrics,":[154],"our":[155],"shows":[157],"sustains":[161],"substantially":[162],"higher":[163],"beyond":[165],"HBM-fit":[167],"regime":[168],"(e.g.,":[169],"\u2265":[170],"100":[171],"k":[172],"tokens)":[173],"by":[174],"reducing":[175],"per-segment":[176],"residual":[177],"stalls":[178],"compared":[179],"while":[183],"avoiding":[184],"severe":[186],"collapse":[187],"re-computation.":[189],"We":[190],"close":[191],"sensitivity":[193],"analyses,":[194],"limitations,":[195],"guidance":[197],"deploying":[199],"fabrics":[201],"long-context":[203],"serving":[204],"stacks.":[205]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-04-03T00:00:00"}