{"id":"https://openalex.org/W4414898558","doi":"https://doi.org/10.1109/cluster59342.2025.11186468","title":"Scalable and Fast Inference Serving via Hybrid Communication Scheduling on Heterogeneous Networks","display_name":"Scalable and Fast Inference Serving via Hybrid Communication Scheduling on Heterogeneous Networks","publication_year":2025,"publication_date":"2025-09-02","ids":{"openalex":"https://openalex.org/W4414898558","doi":"https://doi.org/10.1109/cluster59342.2025.11186468"},"language":"en","primary_location":{"id":"doi:10.1109/cluster59342.2025.11186468","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cluster59342.2025.11186468","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Conference on Cluster Computing (CLUSTER)","raw_type":"proceedings-article"},"type":"article","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/A5000788458","display_name":"Gonglong Chen","orcid":"https://orcid.org/0000-0002-7833-6458"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Gonglong Chen","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology, CAS,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology, CAS,Shenzhen,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086161245","display_name":"Jiamei Lv","orcid":"https://orcid.org/0000-0002-6673-722X"},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Jiamei Lv","raw_affiliation_strings":["Zhejiang University,Hangzhou,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University,Hangzhou,China","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086228898","display_name":"Kejiang Ye","orcid":"https://orcid.org/0000-0001-8985-2792"},"institutions":[{"id":"https://openalex.org/I4210145761","display_name":"Shenzhen Institutes of Advanced Technology","ror":"https://ror.org/04gh4er46","country_code":"CN","type":"facility","lineage":["https://openalex.org/I19820366","https://openalex.org/I4210145761"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Kejiang Ye","raw_affiliation_strings":["Shenzhen Institutes of Advanced Technology, CAS,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Shenzhen Institutes of Advanced Technology, CAS,Shenzhen,China","institution_ids":["https://openalex.org/I4210145761"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5054698524","display_name":"Tao Gu","orcid":"https://orcid.org/0000-0002-1350-6639"},"institutions":[{"id":"https://openalex.org/I99043593","display_name":"Macquarie University","ror":"https://ror.org/01sf06y89","country_code":"AU","type":"education","lineage":["https://openalex.org/I99043593"]}],"countries":["AU"],"is_corresponding":false,"raw_author_name":"Tao Gu","raw_affiliation_strings":["Macquarie University,Sydney,Australia"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Macquarie University,Sydney,Australia","institution_ids":["https://openalex.org/I99043593"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5012773300","display_name":"Chengzhong Xu","orcid":"https://orcid.org/0000-0001-9480-0356"},"institutions":[{"id":"https://openalex.org/I204512498","display_name":"University of Macau","ror":"https://ror.org/01r4q9n85","country_code":"MO","type":"education","lineage":["https://openalex.org/I204512498"]},{"id":"https://openalex.org/I6469544","display_name":"City University of Macau","ror":"https://ror.org/04gpd4q15","country_code":"MO","type":"education","lineage":["https://openalex.org/I6469544"]}],"countries":["MO"],"is_corresponding":false,"raw_author_name":"Chengzhong Xu","raw_affiliation_strings":["University of Macau,Macau SAR,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Macau,Macau SAR,China","institution_ids":["https://openalex.org/I6469544","https://openalex.org/I204512498"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":5,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.2879412,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"12"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10080","display_name":"Energy Efficient Wireless Sensor Networks","score":0.9672999978065491,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T10080","display_name":"Energy Efficient Wireless Sensor Networks","score":0.9672999978065491,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/T11932","display_name":"Wireless Body Area Networks","score":0.9257000088691711,"subfield":{"id":"https://openalex.org/subfields/2204","display_name":"Biomedical 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/scalability","display_name":"Scalability","score":0.7062000036239624},{"id":"https://openalex.org/keywords/server","display_name":"Server","score":0.6050000190734863},{"id":"https://openalex.org/keywords/heterogeneous-network","display_name":"Heterogeneous network","score":0.5733000040054321},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.5654000043869019},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5324000120162964},{"id":"https://openalex.org/keywords/synchronization","display_name":"Synchronization (alternating current)","score":0.5234000086784363},{"id":"https://openalex.org/keywords/testbed","display_name":"Testbed","score":0.5087000131607056},{"id":"https://openalex.org/keywords/overhead","display_name":"Overhead (engineering)","score":0.4503999948501587}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8337000012397766},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7062000036239624},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.650600016117096},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.6050000190734863},{"id":"https://openalex.org/C158207573","wikidata":"https://www.wikidata.org/wiki/Q5747224","display_name":"Heterogeneous network","level":4,"score":0.5733000040054321},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.5654000043869019},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5324000120162964},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.5234000086784363},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.5087000131607056},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.46149998903274536},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.4503999948501587},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.36340001225471497},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.29820001125335693},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.29750001430511475},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.29120001196861267},{"id":"https://openalex.org/C108734733","wikidata":"https://www.wikidata.org/wiki/Q1172333","display_name":"Data synchronization","level":3,"score":0.28299999237060547},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.26660001277923584},{"id":"https://openalex.org/C66882249","wikidata":"https://www.wikidata.org/wiki/Q169336","display_name":"Homogeneous","level":2,"score":0.26460000872612},{"id":"https://openalex.org/C101765175","wikidata":"https://www.wikidata.org/wiki/Q577764","display_name":"Communications system","level":2,"score":0.26030001044273376},{"id":"https://openalex.org/C138959212","wikidata":"https://www.wikidata.org/wiki/Q1806783","display_name":"Load balancing (electrical power)","level":3,"score":0.25619998574256897},{"id":"https://openalex.org/C107568181","wikidata":"https://www.wikidata.org/wiki/Q5319000","display_name":"Dynamic priority scheduling","level":3,"score":0.2549999952316284}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/cluster59342.2025.11186468","is_oa":false,"landing_page_url":"https://doi.org/10.1109/cluster59342.2025.11186468","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE International Conference on Cluster Computing (CLUSTER)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G7672641152","display_name":null,"funder_award_id":"92267105","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":19,"referenced_works":["https://openalex.org/W3185341429","https://openalex.org/W3188270315","https://openalex.org/W3190053701","https://openalex.org/W4317038430","https://openalex.org/W4318540560","https://openalex.org/W4321636575","https://openalex.org/W4327911839","https://openalex.org/W4386260612","https://openalex.org/W4386365337","https://openalex.org/W4387321091","https://openalex.org/W4387760944","https://openalex.org/W4389988990","https://openalex.org/W4389989295","https://openalex.org/W4394892714","https://openalex.org/W4401176521","https://openalex.org/W4401507987","https://openalex.org/W4402671766","https://openalex.org/W4402896821","https://openalex.org/W4407131011"],"related_works":[],"abstract_inverted_index":{"Advances":[0],"in":[1,34,47,123],"large":[2],"language":[3],"models":[4],"(LLMs)":[5],"have":[6],"opened":[7],"up":[8,71],"new":[9,16],"possibilities":[10],"across":[11],"various":[12],"fields,":[13],"fueling":[14],"a":[15,31,129,162,165],"wave":[17],"of":[18,57,65],"interactive":[19],"AI":[20],"applications":[21],"such":[22],"as":[23],"DeepSeek":[24],"and":[25,131,145,157,170],"ChatGPT.":[26],"Inference":[27],"serving":[28,49,113,134],"systems":[29,75],"play":[30],"crucial":[32],"role":[33],"supporting":[35],"these":[36,105],"applications.":[37],"Recent":[38],"research":[39],"indicates":[40],"that":[41,115,177],"when":[42],"crossserver":[43],"parallelization":[44],"is":[45],"enabled":[46],"inference":[48,60,112,133],"systems,":[50],"data":[51,121],"synchronization":[52,81,122],"overhead":[53,67],"can":[54],"exceed":[55],"65%":[56],"the":[58,63],"total":[59],"delay,":[61],"making":[62],"reduction":[64],"communication":[66,146],"essential":[68],"for":[69,99,141,153],"speeding":[70],"inference.":[72],"While":[73],"existing":[74],"accelerate":[76,120],"cross-server":[77],"communications":[78],"by":[79,136,181],"offloading":[80],"operations":[82],"to":[83,119,191],"programmable":[84,172],"switches,":[85],"they":[86],"often":[87],"suffer":[88],"from":[89],"limited":[90],"aggregation":[91],"throughput":[92],"under":[93],"bursty":[94],"traffic":[95,155],"conditions,":[96],"posing":[97],"challenges":[98],"homogeneous":[100],"network":[101],"environments.":[102],"To":[103],"address":[104],"challenges,":[106],"we":[107],"propose":[108],"HeroServe,":[109],"an":[110,138,150],"innovative":[111],"system":[114,135],"leverages":[116],"heterogeneous":[117],"networks":[118],"distributed":[124],"clusters.":[125],"Our":[126],"approach":[127],"enables":[128],"fast":[130],"scalable":[132],"employing":[137],"offline":[139],"planner":[140],"joint":[142],"computation":[143],"allocation":[144],"scheduling,":[147],"along":[148],"with":[149],"online":[151],"scheduler":[152],"dynamic":[154],"management":[156],"load":[158],"balancing.":[159],"We":[160],"implement":[161],"prototype":[163],"on":[164],"testbed":[166],"comprising":[167],"six":[168],"servers":[169],"two":[171],"switches.":[173],"Experimental":[174],"results":[175],"demonstrate":[176],"HeroServe":[178],"improves":[179],"scalability":[180],"<tex":[182],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[183],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">$1.53":[184],"\\times$</tex>":[185],"while":[186],"achieving":[187],"lower":[188],"latency":[189],"compared":[190],"state-of-the-art":[192],"solutions.":[193]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-10-10T00:00:00"}