{"id":"https://openalex.org/W7160905955","doi":"https://doi.org/10.48550/arxiv.2605.10905","title":"TLX: Hardware-Native, Evolvable MIMW GPU Compiler for Large-scale Production Environments","display_name":"TLX: Hardware-Native, Evolvable MIMW GPU Compiler for Large-scale Production Environments","publication_year":2026,"publication_date":"2026-05-11","ids":{"openalex":"https://openalex.org/W7160905955","doi":"https://doi.org/10.48550/arxiv.2605.10905"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.10905","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10905","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"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":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.10905","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135916606","display_name":"Yue Guan","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Guan, Yue","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135944284","display_name":"Hongtao Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yu, Hongtao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135989352","display_name":"Peng Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5069028055","display_name":"Daohang Shi","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shi, Daohang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135994272","display_name":"Karthik Manivannan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Manivannan, Karthik","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135968256","display_name":"Nicholas J Riasanovsky","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Riasanovsky, Nicholas J","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5027121583","display_name":"Manman Ren","orcid":"https://orcid.org/0009-0006-7023-9449"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ren, Manman","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135959054","display_name":"Lei Wang","orcid":"https://orcid.org/0009-0004-1348-8583"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Lei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5079283501","display_name":"Shane Nay","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nay, Shane","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5047566695","display_name":"Partha Kanuparthy","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kanuparthy, Partha","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135978286","display_name":"Zaifeng Pan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pan, Zaifeng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135919100","display_name":"Zhengding Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Zhengding","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135971956","display_name":"Yufei Ding","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ding, Yufei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":13,"corresponding_author_ids":["https://openalex.org/A5135916606"],"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":0.9462000131607056,"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":0.9462000131607056,"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/T10904","display_name":"Embedded Systems Design Techniques","score":0.006800000090152025,"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/T12292","display_name":"Graph Theory and Algorithms","score":0.006599999964237213,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/compiler","display_name":"Compiler","score":0.7985000014305115},{"id":"https://openalex.org/keywords/orchestration","display_name":"Orchestration","score":0.7268000245094299},{"id":"https://openalex.org/keywords/asynchronous-communication","display_name":"Asynchronous communication","score":0.5990999937057495},{"id":"https://openalex.org/keywords/synchronization","display_name":"Synchronization (alternating current)","score":0.5687000155448914},{"id":"https://openalex.org/keywords/granularity","display_name":"Granularity","score":0.5389000177383423},{"id":"https://openalex.org/keywords/execution-model","display_name":"Execution model","score":0.41609999537467957},{"id":"https://openalex.org/keywords/concurrency","display_name":"Concurrency","score":0.39809998869895935},{"id":"https://openalex.org/keywords/data-structure","display_name":"Data structure","score":0.37070000171661377},{"id":"https://openalex.org/keywords/code-generation","display_name":"Code generation","score":0.3684000074863434}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8192999958992004},{"id":"https://openalex.org/C169590947","wikidata":"https://www.wikidata.org/wiki/Q47506","display_name":"Compiler","level":2,"score":0.7985000014305115},{"id":"https://openalex.org/C199168358","wikidata":"https://www.wikidata.org/wiki/Q3367000","display_name":"Orchestration","level":3,"score":0.7268000245094299},{"id":"https://openalex.org/C151319957","wikidata":"https://www.wikidata.org/wiki/Q752739","display_name":"Asynchronous communication","level":2,"score":0.5990999937057495},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.5820000171661377},{"id":"https://openalex.org/C2778562939","wikidata":"https://www.wikidata.org/wiki/Q1298791","display_name":"Synchronization (alternating current)","level":3,"score":0.5687000155448914},{"id":"https://openalex.org/C177774035","wikidata":"https://www.wikidata.org/wiki/Q1246948","display_name":"Granularity","level":2,"score":0.5389000177383423},{"id":"https://openalex.org/C2776834041","wikidata":"https://www.wikidata.org/wiki/Q25346349","display_name":"Execution model","level":2,"score":0.41609999537467957},{"id":"https://openalex.org/C193702766","wikidata":"https://www.wikidata.org/wiki/Q1414548","display_name":"Concurrency","level":2,"score":0.39809998869895935},{"id":"https://openalex.org/C162319229","wikidata":"https://www.wikidata.org/wiki/Q175263","display_name":"Data structure","level":2,"score":0.37070000171661377},{"id":"https://openalex.org/C133162039","wikidata":"https://www.wikidata.org/wiki/Q1061077","display_name":"Code generation","level":3,"score":0.3684000074863434},{"id":"https://openalex.org/C52723943","wikidata":"https://www.wikidata.org/wiki/Q1127410","display_name":"Serialization","level":2,"score":0.3682999908924103},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.3675000071525574},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.36730000376701355},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.36239999532699585},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.35100001096725464},{"id":"https://openalex.org/C160145156","wikidata":"https://www.wikidata.org/wiki/Q778586","display_name":"Executable","level":2,"score":0.3449999988079071},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.3370000123977661},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.33489999175071716},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3319999873638153},{"id":"https://openalex.org/C183003079","wikidata":"https://www.wikidata.org/wiki/Q1000371","display_name":"Personalization","level":2,"score":0.3197000026702881},{"id":"https://openalex.org/C34165917","wikidata":"https://www.wikidata.org/wiki/Q188267","display_name":"Programming paradigm","level":2,"score":0.31299999356269836},{"id":"https://openalex.org/C19024347","wikidata":"https://www.wikidata.org/wiki/Q211496","display_name":"High-level programming language","level":3,"score":0.31029999256134033},{"id":"https://openalex.org/C168065819","wikidata":"https://www.wikidata.org/wiki/Q845566","display_name":"Debugging","level":2,"score":0.3093999922275543},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.29760000109672546},{"id":"https://openalex.org/C115903868","wikidata":"https://www.wikidata.org/wiki/Q80993","display_name":"Software engineering","level":1,"score":0.29190000891685486},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.2858999967575073},{"id":"https://openalex.org/C131531359","wikidata":"https://www.wikidata.org/wiki/Q5421926","display_name":"Extensible programming","level":5,"score":0.2847000062465668},{"id":"https://openalex.org/C61423126","wikidata":"https://www.wikidata.org/wiki/Q187432","display_name":"Scripting language","level":2,"score":0.2736000120639801},{"id":"https://openalex.org/C113843644","wikidata":"https://www.wikidata.org/wiki/Q901882","display_name":"Interface (matter)","level":4,"score":0.26350000500679016},{"id":"https://openalex.org/C2778029271","wikidata":"https://www.wikidata.org/wiki/Q5421931","display_name":"Extension (predicate logic)","level":2,"score":0.2531000077724457},{"id":"https://openalex.org/C139968098","wikidata":"https://www.wikidata.org/wiki/Q3055454","display_name":"Development environment","level":2,"score":0.25110000371932983}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.10905","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10905","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.10905","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10905","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"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":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Modern":[0],"GPUs":[1],"increasingly":[2],"rely":[3],"on":[4,15],"specialized":[5],"hardware":[6,48],"units":[7],"and":[8,21,112,141],"asynchronous":[9,110],"coordination":[10],"mechanisms,":[11],"so":[12],"performance":[13],"depends":[14],"orchestrating":[16],"data":[17],"movement,":[18],"tensor-core":[19],"computation,":[20],"synchronization":[22],"rather":[23],"than":[24],"exposing":[25,102],"more":[26],"thread-level":[27],"parallelism.":[28],"This":[29],"creates":[30],"a":[31],"programming-model":[32],"tension:":[33],"if":[34,50],"too":[35,51],"much":[36,52],"execution":[37],"structure":[38],"is":[39,53,147],"hidden,":[40],"the":[41,55,62],"compiler":[42],"must":[43],"catch":[44],"up":[45],"to":[46,100],"new":[47],"mechanisms;":[49],"exposed,":[54],"burden":[56],"of":[57],"orchestration":[58,78],"falls":[59],"back":[60],"onto":[61],"programmer.":[63],"We":[64],"present":[65],"TLX":[66,92,119],"(Triton":[67],"Low-level":[68],"Language":[69],"Extensions),":[70],"built":[71],"around":[72],"MIMW":[73],"(Multi-Instruction,":[74],"Multi-Warp),":[75],"which":[76],"expresses":[77],"at":[79,150],"warp-group":[80],"granularity":[81],"while":[82,127],"preserving":[83],"Triton's":[84],"productive":[85],"blocked":[86],"programming":[87],"model":[88],"for":[89,105],"regular":[90],"computation.":[91],"realizes":[93],"this":[94],"idea":[95],"as":[96],"an":[97],"embedded":[98],"extension":[99],"Triton,":[101],"explicit":[103],"interfaces":[104],"multi-warp":[106],"execution,":[107],"local-memory":[108],"orchestration,":[109],"operations,":[111],"cluster-aware":[113],"control.":[114],"Our":[115,145],"evaluation":[116],"shows":[117],"that":[118],"supports":[120],"substantial":[121],"customization":[122],"with":[123,130],"limited":[124],"development":[125],"effort":[126],"remaining":[128],"competitive":[129],"state-of-the-art":[131],"implementations.":[132],"TLX-authored":[133],"kernels":[134],"have":[135],"been":[136],"deployed":[137],"in":[138],"large-scale":[139],"training":[140],"inference":[142],"production":[143],"systems.":[144],"code":[146],"open":[148],"sourced":[149],"https://github.com/facebookexperimental/triton.":[151]},"counts_by_year":[],"updated_date":"2026-05-13T06:11:35.469786","created_date":"2026-05-13T00:00:00"}
