{"id":"https://openalex.org/W7225823282","doi":"https://doi.org/10.5281/zenodo.24144821","title":"Anudhi: A Fast Inference Engine for Edge Language Models","display_name":"Anudhi: A Fast Inference Engine for Edge Language Models","publication_year":2026,"publication_date":"2026-11-11","ids":{"openalex":"https://openalex.org/W7225823282","doi":"https://doi.org/10.5281/zenodo.24144821"},"language":"en","primary_location":{"id":"doi:10.5281/zenodo.24144821","doi":"https://doi.org/10.5281/zenodo.24144821","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24144821","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.5281/zenodo.24144821","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5160888945","display_name":"Sayan Chowdhury","orcid":"https://orcid.org/0009-0004-8421-0087","observed_orcids":["https://orcid.org/0009-0004-8421-0087"]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Chowdhury, Sayan","raw_affiliation_strings":["Kalaris Labs"],"raw_orcid":"https://orcid.org/0009-0004-8421-0087","affiliations":[{"raw_affiliation_string":"Kalaris Labs","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5160888945"],"corresponding_institution_ids":[],"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":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.6499000191688538,"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"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.6499000191688538,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.1858000010251999,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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":0.04500000178813934,"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":[],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8429999947547913},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.489300012588501},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.4440999925136566},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.44040000438690186},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.4343000054359436},{"id":"https://openalex.org/C180591056","wikidata":"https://www.wikidata.org/wiki/Q654437","display_name":"Executor","level":2,"score":0.4059000015258789},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.38420000672340393},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.3580999970436096},{"id":"https://openalex.org/C55166926","wikidata":"https://www.wikidata.org/wiki/Q2892946","display_name":"Oracle","level":2,"score":0.3546999990940094},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.34950000047683716},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.33329999446868896},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.3109000027179718},{"id":"https://openalex.org/C2779182362","wikidata":"https://www.wikidata.org/wiki/Q17126187","display_name":"Session (web analytics)","level":2,"score":0.3073999881744385},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.3050999939441681},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.30329999327659607},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.2994999885559082},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.296999990940094},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.2962999939918518},{"id":"https://openalex.org/C162262903","wikidata":"https://www.wikidata.org/wiki/Q343527","display_name":"Allocator","level":2,"score":0.2897999882698059},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.2815000116825104},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.2782000005245209},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.27480000257492065},{"id":"https://openalex.org/C3265923","wikidata":"https://www.wikidata.org/wiki/Q669129","display_name":"Arithmetic underflow","level":2,"score":0.2736999988555908},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.27309998869895935},{"id":"https://openalex.org/C173801870","wikidata":"https://www.wikidata.org/wiki/Q201413","display_name":"Heuristic","level":2,"score":0.27070000767707825},{"id":"https://openalex.org/C133162039","wikidata":"https://www.wikidata.org/wiki/Q1061077","display_name":"Code generation","level":3,"score":0.26649999618530273},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.26080000400543213},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.25450000166893005},{"id":"https://openalex.org/C35525427","wikidata":"https://www.wikidata.org/wiki/Q745881","display_name":"Intrusion detection system","level":2,"score":0.2538999915122986},{"id":"https://openalex.org/C146778888","wikidata":"https://www.wikidata.org/wiki/Q836862","display_name":"Installation","level":2,"score":0.2533000111579895},{"id":"https://openalex.org/C1793878","wikidata":"https://www.wikidata.org/wiki/Q1153762","display_name":"Out-of-order execution","level":2,"score":0.2515999972820282}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.5281/zenodo.24144821","doi":"https://doi.org/10.5281/zenodo.24144821","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24144821","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},{"id":"doi:10.5281/zenodo.24144823","doi":"https://doi.org/10.5281/zenodo.24144823","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24144823","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.5281/zenodo.24144821","doi":"https://doi.org/10.5281/zenodo.24144821","is_oa":true,"landing_page_url":"https://doi.org/10.5281/zenodo.24144821","pdf_url":null,"source":{"id":"https://openalex.org/S4306400562","display_name":"Zenodo (CERN European Organization for Nuclear Research)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I67311998","host_organization_name":"European Organization for Nuclear Research","host_organization_lineage":["https://openalex.org/I67311998"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Running":[0],"a":[1,5,13,38,74,85,93,104,167,174,198,227],"language":[2],"model":[3,69],"on":[4,251,265],"phone":[6],"is":[7,22,59,70,140,284,304],"constrained":[8],"by":[9,24,29,286],"four":[10,66],"things":[11],"that":[12,36,78,108,124,205,222],"datacenter":[14],"runtime":[15],"does":[16],"not":[17,28],"face":[18],"together:":[19],"decode":[20],"speed":[21],"set":[23],"memory":[25,39,95],"bandwidth":[26],"and":[27,49,97,146,149,151,166,197,218,230,269],"arithmetic,":[30],"the":[31,45,50,102,171,179,208,212,215,219,235,238,258,270,279,287],"operating":[32],"system":[33],"kills":[34],"processes":[35],"exceed":[37],"budget,":[40],"sustained":[41],"throughput":[42,283,298],"falls":[43,185],"when":[44,178],"device":[46,103,180,297],"heats":[47],"up,":[48],"platform":[51],"forbids":[52],"generating":[53],"code":[54],"at":[55,194,306],"run":[56],"time.":[57],"Anudhi":[58],"an":[60,98,231,262],"inference":[61],"engine":[62],"designed":[63],"around":[64],"those":[65],"constraints.":[67],"A":[68],"compiled":[71],"offline":[72],"into":[73],"single":[75],".kali":[76],"container":[77],"carries":[79],"its":[80],"weights":[81,127],"in":[82,173,291,300],"consumption":[83],"order,":[84],"flat":[86],"execution":[87],"plan":[88,96,109,239],"with":[89,113,144,157,207],"no":[90,114,296],"control":[91],"flow,":[92],"pre-computed":[94],"Ed25519":[99],"signature.":[100],"On":[101,234],"small":[105],"executor":[106,240,264],"replays":[107],"once":[110],"per":[111,163],"token":[112,267],"allocation":[115],"after":[116],"session":[117],"creation.":[118],"Weight-bearing":[119],"operations":[120],"use":[121],"fused":[122,271],"kernels":[123],"multiply":[125],"4-bit":[126,152],"against":[128,226],"32-bit":[129],"floating-point":[130],"activations":[131],"without":[132],"ever":[133],"materializing":[134],"dequantized":[135],"weights.":[136],"The":[137],"key-value":[138],"cache":[139],"typed,":[141],"mutable":[142],"state":[143],"ring":[145],"paged":[147],"layouts":[148],"8-bit":[150],"block":[153],"formats.":[154],"Speculative":[155],"decoding":[156],"exact":[158],"rejection":[159],"sampling":[160],"raises":[161],"tokens":[162],"weight":[164],"pass,":[165],"thermal":[168],"governor":[169],"degrades":[170],"configuration":[172],"fixed,":[175],"reversible":[176],"order":[177],"throttles.":[181],"Every":[182],"backend":[183,204],"segment":[184],"back":[186],"from":[187],"neural":[188],"accelerator":[189],"to":[190,192,244],"GPU":[191],"CPU":[193,253],"load":[195],"time,":[196],"load-time":[199],"numerical":[200],"check":[201],"refuses":[202],"any":[203],"disagrees":[206],"reference.":[209],"We":[210],"describe":[211],"complete":[213],"design,":[214],"C":[216],"interface,":[217],"verification":[220],"harness":[221],"gates":[223,289],"every":[224,266],"kernel":[225,273],"scalar":[228,280],"reference":[229,236],"independent":[232,263],"oracle.":[233],"host":[237],"reproduces":[241],"oracle":[242],"logits":[243],"1.2":[245],"x":[246],"10^-6":[247],"maximum":[248],"absolute":[249],"error":[250],"both":[252],"tiers,":[254],"greedy":[255],"generation":[256],"through":[257],"public":[259],"interface":[260],"matches":[261],"tested,":[268],"AVX2":[272],"runs":[274],"8.0":[275],"times":[276],"faster":[277],"than":[278],"tier.":[281],"Device-class":[282],"governed":[285],"release":[288],"stated":[290],"Section":[292],"12;":[293],"we":[294],"report":[295],"figures":[299],"this":[301],"paper.":[302],"Code":[303],"available":[305],"https://github.com/KalarisLabs/Anudhi.":[307]},"counts_by_year":[],"updated_date":"2026-11-12T12:28:57.059296","created_date":"2026-11-12T00:00:00"}
