{"id":"https://openalex.org/W7162694432","doi":"https://doi.org/10.1109/isqed69900.2026.11534707","title":"CELLM: Confidential and Efficient Lightweight LLM Inference in TEEs","display_name":"CELLM: Confidential and Efficient Lightweight LLM Inference in TEEs","publication_year":2026,"publication_date":"2026-04-08","ids":{"openalex":"https://openalex.org/W7162694432","doi":"https://doi.org/10.1109/isqed69900.2026.11534707"},"language":null,"primary_location":{"id":"doi:10.1109/isqed69900.2026.11534707","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isqed69900.2026.11534707","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 27th International Symposium on Quality Electronic Design (ISQED)","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/A5137233958","display_name":"Hui Feng","orcid":null},"institutions":[{"id":"https://openalex.org/I156087764","display_name":"University of California, Merced","ror":"https://ror.org/00d9ah105","country_code":"US","type":"education","lineage":["https://openalex.org/I156087764"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Hui Feng","raw_affiliation_strings":["University of California,Electrical Engineering and Computer Science,Merced,CA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Electrical Engineering and Computer Science,Merced,CA,USA","institution_ids":["https://openalex.org/I156087764"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5008744701","display_name":"Ben Dong","orcid":null},"institutions":[{"id":"https://openalex.org/I156087764","display_name":"University of California, Merced","ror":"https://ror.org/00d9ah105","country_code":"US","type":"education","lineage":["https://openalex.org/I156087764"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ben Dong","raw_affiliation_strings":["University of California,Electrical Engineering and Computer Science,Merced,CA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Electrical Engineering and Computer Science,Merced,CA,USA","institution_ids":["https://openalex.org/I156087764"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5137226409","display_name":"Qian Wang","orcid":null},"institutions":[{"id":"https://openalex.org/I156087764","display_name":"University of California, Merced","ror":"https://ror.org/00d9ah105","country_code":"US","type":"education","lineage":["https://openalex.org/I156087764"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qian Wang","raw_affiliation_strings":["University of California,Electrical Engineering and Computer Science,Merced,CA,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of California,Electrical Engineering and Computer Science,Merced,CA,USA","institution_ids":["https://openalex.org/I156087764"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I156087764"],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"8"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11424","display_name":"Security and Verification in Computing","score":0.17139999568462372,"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/T11424","display_name":"Security and Verification in Computing","score":0.17139999568462372,"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/T10764","display_name":"Privacy-Preserving Technologies in Data","score":0.15760000050067902,"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/T11241","display_name":"Advanced Malware Detection Techniques","score":0.08160000294446945,"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/confidentiality","display_name":"Confidentiality","score":0.4952000081539154},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.36390000581741333},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.31029999256134033},{"id":"https://openalex.org/keywords/automation","display_name":"Automation","score":0.24300000071525574}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6179999709129333},{"id":"https://openalex.org/C71745522","wikidata":"https://www.wikidata.org/wiki/Q2476929","display_name":"Confidentiality","level":2,"score":0.4952000081539154},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.36390000581741333},{"id":"https://openalex.org/C38652104","wikidata":"https://www.wikidata.org/wiki/Q3510521","display_name":"Computer security","level":1,"score":0.35339999198913574},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34610000252723694},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.31029999256134033},{"id":"https://openalex.org/C127413603","wikidata":"https://www.wikidata.org/wiki/Q11023","display_name":"Engineering","level":0,"score":0.2547000050544739},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.24300000071525574},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.23759999871253967},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.23649999499320984}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/isqed69900.2026.11534707","is_oa":false,"landing_page_url":"https://doi.org/10.1109/isqed69900.2026.11534707","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 27th International Symposium on Quality Electronic Design (ISQED)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W2979832172","https://openalex.org/W3016075089","https://openalex.org/W3096785379","https://openalex.org/W3097924569","https://openalex.org/W3112592596","https://openalex.org/W3170572542","https://openalex.org/W4385679786","https://openalex.org/W4389166734","https://openalex.org/W4392353733","https://openalex.org/W4392806465","https://openalex.org/W4402264420","https://openalex.org/W4405544354","https://openalex.org/W4409362915","https://openalex.org/W4411486597","https://openalex.org/W4412945423","https://openalex.org/W4414197541","https://openalex.org/W4415799217","https://openalex.org/W4416429926","https://openalex.org/W7103748828"],"related_works":[],"abstract_inverted_index":{"Large":[0],"Language":[1],"Models":[2],"(LLMs)":[3],"are":[4],"widely":[5],"used":[6],"in":[7,59,102],"cloud":[8],"environments":[9,61],"but":[10],"face":[11],"serious":[12],"risks":[13],"to":[14,76,98,113,121,134],"model":[15],"confidentiality":[16],"and":[17,46,80,90,139],"data":[18,36],"privacy.":[19],"Trusted":[20],"Execution":[21],"Environments":[22],"(TEEs)":[23],"provide":[24],"confidential":[25],"c":[26],"omputing":[27],"t":[28],"hrough":[29],"s":[30],"trong":[31],"i":[32],"solation":[33],"a":[34],"nd":[35],"encryption,":[37],"yet":[38],"most":[39],"TEEs":[40],"cannot":[41],"meet":[42],"the":[43,70],"heavy":[44],"computation":[45],"memory":[47,71],"demands":[48],"of":[49,73],"LLM":[50,74],"inference.":[51],"In":[52,128],"this":[53],"work,":[54],"we":[55],"evaluate":[56],"lightweight":[57,110],"LLMs":[58,111],"TEE-enabled":[60],"using":[62],"Intel":[63],"Trust":[64],"Domain":[65],"Extensions":[66],"(TDX).":[67],"We":[68,93],"analyze":[69],"usage":[72],"models":[75],"guide":[77],"private-memory":[78],"allocation":[79],"compare":[81],"tokens":[82],"per":[83],"second":[84],"(TPS)":[85],"across":[86],"CPU-only,":[87],"CPU-GPU":[88],"hybrid,":[89],"TEE-based":[91],"settings.":[92],"also":[94],"apply":[95],"quantization":[96,131],"techniques":[97],"further":[99],"accelerate":[100],"inference":[101],"TDX":[103,118],"environment.":[104],"Our":[105],"results":[106],"show":[107],"that":[108],"for":[109],"(up":[112],"7B":[114],"parameters),":[115],"TPS":[116],"on":[117,126],"is":[119],"up":[120,133],"$4":[122],"\\times$":[123,136],"higher":[124,137],"than":[125],"CPU.":[127],"addition,":[129],"INT4":[130],"provides":[132],"$3":[135],"throughput":[138],"reduces":[140],"storage":[141],"by":[142],"approximately":[143],"70%":[144],"compared":[145],"with":[146],"FP16.":[147]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-05-29T00:00:00"}