{"id":"https://openalex.org/W7128438836","doi":"https://doi.org/10.1145/3794845","title":"A Survey on Inference Optimization Techniques for Mixture of Experts Models","display_name":"A Survey on Inference Optimization Techniques for Mixture of Experts Models","publication_year":2026,"publication_date":"2026-02-09","ids":{"openalex":"https://openalex.org/W7128438836","doi":"https://doi.org/10.1145/3794845"},"language":"en","primary_location":{"id":"doi:10.1145/3794845","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3794845","pdf_url":null,"source":{"id":"https://openalex.org/S157921468","display_name":"ACM Computing Surveys","issn_l":"0360-0300","issn":["0360-0300","1557-7341"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Computing Surveys","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1145/3794845","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5125469774","display_name":"Jiacheng Liu","orcid":null},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]},{"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"]},{"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":["CN","HK"],"is_corresponding":true,"raw_author_name":"Jiacheng Liu","raw_affiliation_strings":["Hong Kong University of Science and Technology","Shanghai Jiao Tong University","The Chinese University of Hong Kong"],"raw_orcid":"https://orcid.org/0000-0003-0378-2311","affiliations":[{"raw_affiliation_string":"Hong Kong University of Science and Technology","institution_ids":["https://openalex.org/I200769079"]},{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]},{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125447844","display_name":"Peng Tang","orcid":null},"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":"Peng Tang","raw_affiliation_strings":["Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0009-0005-8196-3953","affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125440555","display_name":"Wenfeng Wang","orcid":null},"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":"Wenfeng Wang","raw_affiliation_strings":["Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0009-0002-8087-6135","affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5123416303","display_name":"Yuhang Ren","orcid":null},"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":"Yuhang Ren","raw_affiliation_strings":["Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0009-0002-5083-4460","affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073267234","display_name":"Xiaofeng Hou","orcid":"https://orcid.org/0000-0003-4372-7851"},"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":"Xiaofeng Hou","raw_affiliation_strings":["Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0000-0003-4372-7851","affiliations":[{"raw_affiliation_string":"Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032708386","display_name":"Pheng\u2010Ann Heng","orcid":"https://orcid.org/0000-0003-3055-5034"},"institutions":[{"id":"https://openalex.org/I177725633","display_name":"Chinese University of Hong Kong","ror":"https://ror.org/00t33hh48","country_code":"HK","type":"education","lineage":["https://openalex.org/I177725633"]}],"countries":["HK"],"is_corresponding":false,"raw_author_name":"Pheng Ann Heng","raw_affiliation_strings":["The Chinese University of Hong Kong"],"raw_orcid":"https://orcid.org/0000-0003-3055-5034","affiliations":[{"raw_affiliation_string":"The Chinese University of Hong Kong","institution_ids":["https://openalex.org/I177725633"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125434187","display_name":"Minyi Guo","orcid":null},"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":"Minyi Guo","raw_affiliation_strings":["Department of Computer Science and Engineering, Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0000-0003-0034-2302","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]},{"author_position":"last","author":{"id":null,"display_name":"Chao Li","orcid":"https://orcid.org/0000-0001-6218-4659"},"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":"Chao Li","raw_affiliation_strings":["Department of Computer Science and Engineering, Shanghai Jiao Tong University"],"raw_orcid":"https://orcid.org/0000-0001-6218-4659","affiliations":[{"raw_affiliation_string":"Department of Computer Science and Engineering, Shanghai Jiao Tong University","institution_ids":["https://openalex.org/I183067930"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5125469774"],"corresponding_institution_ids":["https://openalex.org/I177725633","https://openalex.org/I183067930","https://openalex.org/I200769079"],"apc_list":null,"apc_paid":null,"fwci":29.2147,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.98706977,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":98},"biblio":{"volume":"58","issue":"10","first_page":"1","last_page":"37"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.3156000077724457,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T11704","display_name":"Mobile Crowdsensing and Crowdsourcing","score":0.3156000077724457,"subfield":{"id":"https://openalex.org/subfields/1706","display_name":"Computer Science Applications"},"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/T12101","display_name":"Advanced Bandit Algorithms Research","score":0.09109999984502792,"subfield":{"id":"https://openalex.org/subfields/1803","display_name":"Management Science and Operations Research"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T12814","display_name":"Gaussian Processes and Bayesian Inference","score":0.042899999767541885,"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/inference","display_name":"Inference","score":0.7347000241279602},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.6258000135421753},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.5094000101089478},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.45989999175071716},{"id":"https://openalex.org/keywords/efficient-energy-use","display_name":"Efficient energy use","score":0.4016999900341034},{"id":"https://openalex.org/keywords/cloud-computing","display_name":"Cloud computing","score":0.34150001406669617}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8741999864578247},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7347000241279602},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.6258000135421753},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.5094000101089478},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.46230000257492065},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.45989999175071716},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4262999892234802},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.4016999900341034},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.34150001406669617},{"id":"https://openalex.org/C137836250","wikidata":"https://www.wikidata.org/wiki/Q984063","display_name":"Optimization problem","level":2,"score":0.33709999918937683},{"id":"https://openalex.org/C46743427","wikidata":"https://www.wikidata.org/wiki/Q1341685","display_name":"Inference engine","level":3,"score":0.3343000113964081},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.3255999982357025},{"id":"https://openalex.org/C58328972","wikidata":"https://www.wikidata.org/wiki/Q184609","display_name":"Expert system","level":2,"score":0.3199999928474426},{"id":"https://openalex.org/C155846161","wikidata":"https://www.wikidata.org/wiki/Q1143367","display_name":"Graphical model","level":2,"score":0.3093999922275543},{"id":"https://openalex.org/C105002631","wikidata":"https://www.wikidata.org/wiki/Q4833645","display_name":"Subject-matter expert","level":3,"score":0.28380000591278076},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2750999927520752},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.259799987077713},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.25600001215934753}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3794845","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3794845","pdf_url":null,"source":{"id":"https://openalex.org/S157921468","display_name":"ACM Computing Surveys","issn_l":"0360-0300","issn":["0360-0300","1557-7341"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Computing Surveys","raw_type":"journal-article"}],"best_oa_location":{"id":"doi:10.1145/3794845","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3794845","pdf_url":null,"source":{"id":"https://openalex.org/S157921468","display_name":"ACM Computing Surveys","issn_l":"0360-0300","issn":["0360-0300","1557-7341"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319798","host_organization_name":"Association for Computing Machinery","host_organization_lineage":["https://openalex.org/P4310319798"],"host_organization_lineage_names":["Association for Computing Machinery"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"ACM Computing Surveys","raw_type":"journal-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/7","score":0.8914903998374939,"display_name":"Affordable and clean energy"}],"awards":[{"id":"https://openalex.org/G900233110","display_name":null,"funder_award_id":"62441225 and U24A20234","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":26,"referenced_works":["https://openalex.org/W2092845679","https://openalex.org/W2150884987","https://openalex.org/W2889787757","https://openalex.org/W2963748441","https://openalex.org/W3034296505","https://openalex.org/W3037315640","https://openalex.org/W3175111539","https://openalex.org/W3194676777","https://openalex.org/W4210357113","https://openalex.org/W4384211302","https://openalex.org/W4385573095","https://openalex.org/W4389524452","https://openalex.org/W4394564255","https://openalex.org/W4395007039","https://openalex.org/W4396817294","https://openalex.org/W4399043344","https://openalex.org/W4401832716","https://openalex.org/W4402671991","https://openalex.org/W4403006781","https://openalex.org/W4404792840","https://openalex.org/W4408953254","https://openalex.org/W4409131852","https://openalex.org/W4412886668","https://openalex.org/W4412888953","https://openalex.org/W4415797136","https://openalex.org/W4415798217"],"related_works":[],"abstract_inverted_index":{"The":[0],"emergence":[1],"of":[2,5,153,175],"large-scale":[3],"Mixture":[4],"Experts":[6],"(MoE)":[7],"models":[8,33,52],"represents":[9],"a":[10,61,150,186],"significant":[11,35],"advancement":[12],"in":[13,37,164,178],"artificial":[14],"intelligence,":[15],"offering":[16],"larger":[17],"model":[18,76],"capacity":[19],"and":[20,28,41,71,95,107,123,137,143,156,160,172],"computational":[21,38],"efficiency":[22],"through":[23],"conditional":[24],"computation.":[25],"However,":[26],"deploying":[27],"running":[29],"inference":[30,166,180],"on":[31],"these":[32],"presents":[34],"challenges":[36,159],"resources,":[39],"latency,":[40],"energy":[42,144],"efficiency.":[43,145],"This":[44,146],"comprehensive":[45],"survey":[46,147],"analyzes":[47],"optimization":[48,66,181],"techniques":[49,90],"for":[50],"MoE":[51,165,179],"across":[53],"the":[54,75,112,173],"entire":[55],"system":[56,113],"stack.":[57],"We":[58],"first":[59],"establish":[60],"taxonomical":[62],"framework":[63],"that":[64,127,140],"categorizes":[65],"approaches":[67],"into":[68,134],"model-level,":[69],"system-level,":[70],"hardware-level":[72],"optimizations.":[73],"At":[74,111],"level,":[77,114],"we":[78,115,132,183],"examine":[79],"architectural":[80],"innovations":[81],"including":[82,103],"efficient":[83,124],"expert":[84,108],"design,":[85],"attention":[86],"mechanisms,":[87,122],"various":[88],"compression":[89],"such":[91],"as":[92,98,100],"pruning,":[93],"quantization,":[94],"knowledge":[96],"distillation,":[97],"well":[99],"algorithm":[101],"improvement":[102],"dynamic":[104],"routing":[105],"strategies":[106,139],"merging":[109],"methods.":[110],"investigate":[116],"distributed":[117],"computing":[118],"approaches,":[119],"load":[120],"balancing":[121],"scheduling":[125],"algorithms":[126],"enable":[128],"scalable":[129],"deployment.":[130],"Furthermore,":[131],"delve":[133],"hardware-specific":[135],"optimizations":[136],"co-design":[138],"maximize":[141],"throughput":[142],"provides":[148],"both":[149],"structured":[151],"overview":[152],"existing":[154],"solutions":[155],"identifies":[157],"key":[158],"promising":[161],"research":[162],"directions":[163],"optimization.":[167],"To":[168],"facilitate":[169],"ongoing":[170],"updates":[171],"sharing":[174],"cutting-edge":[176],"advances":[177],"research,":[182],"have":[184],"established":[185],"repository":[187],"accessible":[188],"at":[189],"https://github.com/MoE-Inf/awesome-moe-inference/":[190],".":[191]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":1}],"updated_date":"2026-05-21T06:26:12.895304","created_date":"2026-02-10T00:00:00"}