{"id":"https://openalex.org/W7162798286","doi":"https://doi.org/10.1145/3805653","title":"WISP: Waste- and Interference-Suppressed Distributed Speculative LLM Serving at the Edge via Dynamic Drafting and SLO-Aware Batching","display_name":"WISP: Waste- and Interference-Suppressed Distributed Speculative LLM Serving at the Edge via Dynamic Drafting and SLO-Aware Batching","publication_year":2026,"publication_date":"2026-05-29","ids":{"openalex":"https://openalex.org/W7162798286","doi":"https://doi.org/10.1145/3805653"},"language":"en","primary_location":{"id":"doi:10.1145/3805653","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3805653","pdf_url":null,"source":{"id":"https://openalex.org/S4210193547","display_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","issn_l":"2476-1249","issn":["2476-1249"],"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":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://pureadmin.qub.ac.uk/ws/files/690573404/3805653.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137334781","display_name":"Xiangchen Li","orcid":"https://orcid.org/0000-0002-5293-5348"},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiangchen Li","raw_affiliation_strings":["Virginia Tech, Blacksburg, VA, USA"],"raw_orcid":"https://orcid.org/0000-0002-5293-5348","affiliations":[{"raw_affiliation_string":"Virginia Tech, Blacksburg, VA, USA","institution_ids":["https://openalex.org/I859038795"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125658018","display_name":"Jiakun Fan","orcid":null},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiakun Fan","raw_affiliation_strings":["Virginia Tech, Blacksburg, VA, USA"],"raw_orcid":"https://orcid.org/0009-0004-4563-8697","affiliations":[{"raw_affiliation_string":"Virginia Tech, Blacksburg, VA, USA","institution_ids":["https://openalex.org/I859038795"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137370352","display_name":"Qingyuan Wang","orcid":"https://orcid.org/0000-0002-7879-4328"},"institutions":[{"id":"https://openalex.org/I100930933","display_name":"University College Dublin","ror":"https://ror.org/05m7pjf47","country_code":"IE","type":"education","lineage":["https://openalex.org/I100930933"]}],"countries":["IE"],"is_corresponding":false,"raw_author_name":"Qingyuan Wang","raw_affiliation_strings":["University College Dublin, Dublin, Ireland"],"raw_orcid":"https://orcid.org/0000-0002-7879-4328","affiliations":[{"raw_affiliation_string":"University College Dublin, Dublin, Ireland","institution_ids":["https://openalex.org/I100930933"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005011713","display_name":"Dimitrios Spatharakis","orcid":"https://orcid.org/0000-0002-0157-7173"},"institutions":[{"id":"https://openalex.org/I174458059","display_name":"National Technical University of Athens","ror":"https://ror.org/03cx6bg69","country_code":"GR","type":"education","lineage":["https://openalex.org/I174458059"]}],"countries":["GR"],"is_corresponding":false,"raw_author_name":"Dimitrios Spatharakis","raw_affiliation_strings":["National Technical University of Athens, Athens, Greece"],"raw_orcid":"https://orcid.org/0000-0002-0157-7173","affiliations":[{"raw_affiliation_string":"National Technical University of Athens, Athens, Greece","institution_ids":["https://openalex.org/I174458059"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5019689955","display_name":"Saeid Ghafouri","orcid":"https://orcid.org/0000-0003-3799-5702"},"institutions":[{"id":"https://openalex.org/I126231945","display_name":"Queen's University Belfast","ror":"https://ror.org/00hswnk62","country_code":"GB","type":"education","lineage":["https://openalex.org/I126231945"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Saeid Ghafouri","raw_affiliation_strings":["Queen's University Belfast, Belfast, United Kingdom"],"raw_orcid":"https://orcid.org/0000-0003-3799-5702","affiliations":[{"raw_affiliation_string":"Queen's University Belfast, Belfast, United Kingdom","institution_ids":["https://openalex.org/I126231945"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5086758775","display_name":"Hans Vandierendonck","orcid":"https://orcid.org/0000-0001-5868-9259"},"institutions":[{"id":"https://openalex.org/I126231945","display_name":"Queen's University Belfast","ror":"https://ror.org/00hswnk62","country_code":"GB","type":"education","lineage":["https://openalex.org/I126231945"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Hans Vandierendonck","raw_affiliation_strings":["Queen's University Belfast, Belfast, United Kingdom"],"raw_orcid":"https://orcid.org/0000-0001-5868-9259","affiliations":[{"raw_affiliation_string":"Queen's University Belfast, Belfast, United Kingdom","institution_ids":["https://openalex.org/I126231945"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5029098934","display_name":"Deepu John","orcid":"https://orcid.org/0000-0002-6139-1100"},"institutions":[{"id":"https://openalex.org/I100930933","display_name":"University College Dublin","ror":"https://ror.org/05m7pjf47","country_code":"IE","type":"education","lineage":["https://openalex.org/I100930933"]}],"countries":["IE"],"is_corresponding":false,"raw_author_name":"Deepu John","raw_affiliation_strings":["University College Dublin, Dublin, Ireland"],"raw_orcid":"https://orcid.org/0000-0002-6139-1100","affiliations":[{"raw_affiliation_string":"University College Dublin, Dublin, Ireland","institution_ids":["https://openalex.org/I100930933"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137360993","display_name":"Bo Ji","orcid":"https://orcid.org/0000-0003-0149-7509"},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bo Ji","raw_affiliation_strings":["Virginia Tech, Blacksburg, USA"],"raw_orcid":"https://orcid.org/0000-0003-0149-7509","affiliations":[{"raw_affiliation_string":"Virginia Tech, Blacksburg, USA","institution_ids":["https://openalex.org/I859038795"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137343757","display_name":"Ali R. Butt","orcid":"https://orcid.org/0000-0002-0871-7263"},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ali R. Butt","raw_affiliation_strings":["Virginia Tech, Blacksburg, USA"],"raw_orcid":"https://orcid.org/0000-0002-0871-7263","affiliations":[{"raw_affiliation_string":"Virginia Tech, Blacksburg, USA","institution_ids":["https://openalex.org/I859038795"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5113899214","display_name":"Dimitrios Nikolopoulos","orcid":null},"institutions":[{"id":"https://openalex.org/I859038795","display_name":"Virginia Tech","ror":"https://ror.org/02smfhw86","country_code":"US","type":"education","lineage":["https://openalex.org/I859038795"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dimitrios Nikolopoulos","raw_affiliation_strings":["Virginia Tech, Blacksburg, USA"],"raw_orcid":"https://orcid.org/0000-0003-0217-8307","affiliations":[{"raw_affiliation_string":"Virginia Tech, Blacksburg, USA","institution_ids":["https://openalex.org/I859038795"]}]}],"institutions":[],"countries_distinct_count":4,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.76225207,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"10","issue":"2","first_page":"1","last_page":"31"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T14347","display_name":"Big Data and Digital Economy","score":0.17080000042915344,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T14347","display_name":"Big Data and Digital Economy","score":0.17080000042915344,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T10181","display_name":"Natural Language Processing Techniques","score":0.09600000083446503,"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/T11948","display_name":"Machine Learning in Materials Science","score":0.08789999783039093,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/goodput","display_name":"Goodput","score":0.8421000242233276},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.7324000000953674},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.51419997215271},{"id":"https://openalex.org/keywords/workload","display_name":"Workload","score":0.5099999904632568},{"id":"https://openalex.org/keywords/edge-device","display_name":"Edge device","score":0.49869999289512634},{"id":"https://openalex.org/keywords/scheduling","display_name":"Scheduling (production processes)","score":0.4893999993801117},{"id":"https://openalex.org/keywords/enhanced-data-rates-for-gsm-evolution","display_name":"Enhanced Data Rates for GSM Evolution","score":0.48579999804496765},{"id":"https://openalex.org/keywords/single-point-of-failure","display_name":"Single point of failure","score":0.4300000071525574},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.4189999997615814},{"id":"https://openalex.org/keywords/process","display_name":"Process (computing)","score":0.4056999981403351}],"concepts":[{"id":"https://openalex.org/C94022561","wikidata":"https://www.wikidata.org/wiki/Q1172393","display_name":"Goodput","level":4,"score":0.8421000242233276},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8371999859809875},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.7324000000953674},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.6360999941825867},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.51419997215271},{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.5099999904632568},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.49869999289512634},{"id":"https://openalex.org/C206729178","wikidata":"https://www.wikidata.org/wiki/Q2271896","display_name":"Scheduling (production processes)","level":2,"score":0.4893999993801117},{"id":"https://openalex.org/C162307627","wikidata":"https://www.wikidata.org/wiki/Q204833","display_name":"Enhanced Data Rates for GSM Evolution","level":2,"score":0.48579999804496765},{"id":"https://openalex.org/C165136773","wikidata":"https://www.wikidata.org/wiki/Q1363179","display_name":"Single point of failure","level":2,"score":0.4300000071525574},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.4189999997615814},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4056999981403351},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.397599995136261},{"id":"https://openalex.org/C2778869765","wikidata":"https://www.wikidata.org/wiki/Q6028363","display_name":"Inefficiency","level":2,"score":0.38929998874664307},{"id":"https://openalex.org/C162932704","wikidata":"https://www.wikidata.org/wiki/Q1058791","display_name":"Synchronizing","level":3,"score":0.37959998846054077},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.37389999628067017},{"id":"https://openalex.org/C2778456923","wikidata":"https://www.wikidata.org/wiki/Q5337692","display_name":"Edge computing","level":3,"score":0.36309999227523804},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.35929998755455017},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.3346000015735626},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3260999917984009},{"id":"https://openalex.org/C113954288","wikidata":"https://www.wikidata.org/wiki/Q186885","display_name":"Timestamp","level":2,"score":0.3158999979496002},{"id":"https://openalex.org/C2776493592","wikidata":"https://www.wikidata.org/wiki/Q5158717","display_name":"Succinctness","level":2,"score":0.3111000061035156},{"id":"https://openalex.org/C2780945871","wikidata":"https://www.wikidata.org/wiki/Q194274","display_name":"Backup","level":2,"score":0.3070000112056732},{"id":"https://openalex.org/C81081738","wikidata":"https://www.wikidata.org/wiki/Q55542","display_name":"Lossless compression","level":3,"score":0.30250000953674316},{"id":"https://openalex.org/C75949130","wikidata":"https://www.wikidata.org/wiki/Q848010","display_name":"Database transaction","level":2,"score":0.30000001192092896},{"id":"https://openalex.org/C2780310539","wikidata":"https://www.wikidata.org/wiki/Q12547192","display_name":"Imperfect","level":2,"score":0.29190000891685486},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.2912999987602234},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2897999882698059},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.2797999978065491},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.27489998936653137},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.27239999175071716},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.26510000228881836},{"id":"https://openalex.org/C55439883","wikidata":"https://www.wikidata.org/wiki/Q360812","display_name":"Correctness","level":2,"score":0.25209999084472656}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3805653","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3805653","pdf_url":null,"source":{"id":"https://openalex.org/S4210193547","display_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","issn_l":"2476-1249","issn":["2476-1249"],"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":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the ACM on Measurement and Analysis of Computing Systems","raw_type":"journal-article"},{"id":"pmh:oai:pure.qub.ac.uk/portal:publications/257619a9-b76f-40ee-8a52-aa6339e3a337","is_oa":true,"landing_page_url":"https://pure.qub.ac.uk/en/publications/257619a9-b76f-40ee-8a52-aa6339e3a337","pdf_url":"https://pureadmin.qub.ac.uk/ws/files/690573404/3805653.pdf","source":{"id":"https://openalex.org/S4306402319","display_name":"Research Portal (Queen's University Belfast)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I126231945","host_organization_name":"Queen's University Belfast","host_organization_lineage":["https://openalex.org/I126231945"],"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":"Li, X, Fan, J, Wang, Q, Spatharakis, D, Ghafouri, S, Vandierendonck, H, John, D, Ji, B, Butt, A R & Nikolopoulos, D 2026, 'WISP: waste- and interference-suppressed distributed speculative LLM serving at the edge via dynamic drafting and SLO-aware batching', Proceedings of the ACM on Measurement and Analysis of Computing Systems, vol. 10, no. 2, 55. https://doi.org/10.1145/3805653","raw_type":"info:eu-repo/semantics/article"}],"best_oa_location":{"id":"pmh:oai:pure.qub.ac.uk/portal:publications/257619a9-b76f-40ee-8a52-aa6339e3a337","is_oa":true,"landing_page_url":"https://pure.qub.ac.uk/en/publications/257619a9-b76f-40ee-8a52-aa6339e3a337","pdf_url":"https://pureadmin.qub.ac.uk/ws/files/690573404/3805653.pdf","source":{"id":"https://openalex.org/S4306402319","display_name":"Research Portal (Queen's University Belfast)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I126231945","host_organization_name":"Queen's University Belfast","host_organization_lineage":["https://openalex.org/I126231945"],"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":"Li, X, Fan, J, Wang, Q, Spatharakis, D, Ghafouri, S, Vandierendonck, H, John, D, Ji, B, Butt, A R & Nikolopoulos, D 2026, 'WISP: waste- and interference-suppressed distributed speculative LLM serving at the edge via dynamic drafting and SLO-aware batching', Proceedings of the ACM on Measurement and Analysis of Computing Systems, vol. 10, no. 2, 55. https://doi.org/10.1145/3805653","raw_type":"info:eu-repo/semantics/article"},"sustainable_development_goals":[{"display_name":"Decent work and economic growth","id":"https://metadata.un.org/sdg/8","score":0.4971916675567627}],"awards":[{"id":"https://openalex.org/G3294571139","display_name":"U.S.-Ireland R&D Partnership:CNS:Small:SWEET: Hardware and Software for Sustainable Wearable Edge Intelligence","funder_award_id":"2315851","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7162798286.pdf","grobid_xml":"https://content.openalex.org/works/W7162798286.grobid-xml"},"referenced_works_count":10,"referenced_works":["https://openalex.org/W2109488193","https://openalex.org/W2966983573","https://openalex.org/W3186289964","https://openalex.org/W4384520874","https://openalex.org/W4401211704","https://openalex.org/W4402671659","https://openalex.org/W4404134026","https://openalex.org/W4404401018","https://openalex.org/W4416927400","https://openalex.org/W4416955988"],"related_works":[],"abstract_inverted_index":{"As":[0],"Large":[1],"Language":[2],"Models":[3],"(LLMs)":[4],"become":[5],"increasingly":[6],"accessible":[7],"to":[8,50,169,179,184],"end":[9],"users,":[10],"an":[11,120,131],"ever-growing":[12],"number":[13],"of":[14,101,130],"inference":[15,65,126],"requests":[16],"are":[17],"initiated":[18],"from":[19],"edge":[20,44,60,76],"devices":[21,45,61],"and":[22,53,77,90,99,109,122,139,150,171,173,181,187],"computed":[23],"on":[24,40,155],"centralized":[25,185],"GPU":[26],"clusters.":[27],"However,":[28],"the":[29,57,63,72,75,78,97,156],"resulting":[30],"exponential":[31],"growth":[32],"in":[33],"computation":[34],"workload":[35,73],"is":[36],"placing":[37],"significant":[38],"strain":[39],"data":[41],"centers,":[42],"while":[43,80],"remain":[46],"largely":[47],"underutilized,":[48],"leading":[49],"imbalanced":[51],"workloads":[52],"resource":[54],"inefficiency":[55],"across":[56],"network.":[58],"Integrating":[59],"into":[62],"LLM":[64,104,125],"process":[66],"via":[67],"speculative":[68,103],"decoding":[69],"helps":[70],"balance":[71],"between":[74],"cloud,":[79],"maintaining":[81],"lossless":[82],"prediction":[83],"accuracy.":[84],"In":[85],"this":[86],"paper,":[87],"we":[88,116],"identify":[89],"formalize":[91],"two":[92],"critical":[93],"bottlenecks":[94],"that":[95,128,162],"limit":[96],"efficiency":[98,149],"scalability":[100],"distributed":[102,124],"serving:":[105],"Wasted":[106],"Drafting":[107],"Time":[108],"Verification":[110],"Interference.":[111],"To":[112],"address":[113],"these":[114],"challenges,":[115],"propose":[117],"WISP":[118,163],",":[119,189],"efficient":[121],"SLO-aware":[123],"system":[127,165,175],"consists":[129],"intelligent":[132],"speculation":[133],"controller,":[134],"a":[135,140],"verification":[136,141,152],"time":[137],"estimator,":[138],"batch":[142],"scheduler.":[143],"These":[144],"components":[145],"collaboratively":[146],"enhance":[147],"drafting":[148],"optimize":[151],"request":[153],"scheduling":[154],"server.":[157],"Extensive":[158],"numerical":[159],"results":[160],"show":[161],"improves":[164],"capacity":[166],"by":[167,177],"up":[168,178],"2.1\u00d7":[170],"4.1\u00d7,":[172],"increases":[174],"goodput":[176],"1.94\u00d7":[180],"3.7\u00d7,":[182],"compared":[183],"serving":[186],"SLED":[188],"respectively.":[190]},"counts_by_year":[],"updated_date":"2026-07-02T09:51:11.867554","created_date":"2026-05-30T00:00:00"}