{"id":"https://openalex.org/W7106138757","doi":"https://doi.org/10.48550/arxiv.2511.13940","title":"ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels","display_name":"ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels","publication_year":2025,"publication_date":"2025-11-17","ids":{"openalex":"https://openalex.org/W7106138757","doi":"https://doi.org/10.48550/arxiv.2511.13940"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2511.13940","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.13940","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2511.13940","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":null,"display_name":"Sul, Stuart H.","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sul, Stuart H.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Arora, Simran","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Arora, Simran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":null,"display_name":"Spector, Benjamin F.","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Spector, Benjamin F.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"R\u00e9, Christopher","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"R\u00e9, Christopher","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"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":0.9358999729156494,"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":0.9358999729156494,"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/T14347","display_name":"Big Data and Digital Economy","score":0.014299999922513962,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.008299999870359898,"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/bottleneck","display_name":"Bottleneck","score":0.8213000297546387},{"id":"https://openalex.org/keywords/kernel","display_name":"Kernel (algebra)","score":0.6455000042915344},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5921000242233276},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.5785999894142151},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.517300009727478},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.4203999936580658},{"id":"https://openalex.org/keywords/core","display_name":"Core (optical fiber)","score":0.38679999113082886}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8223999738693237},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.8213000297546387},{"id":"https://openalex.org/C74193536","wikidata":"https://www.wikidata.org/wiki/Q574844","display_name":"Kernel (algebra)","level":2,"score":0.6455000042915344},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5921000242233276},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.5785999894142151},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.517300009727478},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.4203999936580658},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.4034999907016754},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.38679999113082886},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.38370001316070557},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.37779998779296875},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.3659999966621399},{"id":"https://openalex.org/C34165917","wikidata":"https://www.wikidata.org/wiki/Q188267","display_name":"Programming paradigm","level":2,"score":0.3481000065803528},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.34450000524520874},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.34130001068115234},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3246999979019165},{"id":"https://openalex.org/C90329073","wikidata":"https://www.wikidata.org/wiki/Q914232","display_name":"Ask price","level":2,"score":0.3221000134944916},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.296999990940094},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.289000004529953},{"id":"https://openalex.org/C123745756","wikidata":"https://www.wikidata.org/wiki/Q1665949","display_name":"Interconnection","level":2,"score":0.2800999879837036},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2671999931335449}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2511.13940","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.13940","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2511.13940","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2511.13940","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"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":{"Inter-GPU":[0],"communication":[1],"has":[2],"become":[3],"a":[4,53,73,104,110],"major":[5],"bottleneck":[6],"for":[7,150,157,163],"modern":[8],"AI":[9],"workloads":[10,42],"as":[11],"models":[12],"scale":[13],"and":[14,43,91,103,123,132,152,160],"improvements":[15,21],"in":[16,22],"hardware":[17],"compute":[18],"throughput":[19],"outpace":[20],"interconnect":[23],"bandwidth.":[24],"Existing":[25],"systems":[26],"mitigate":[27],"this":[28],"through":[29,99],"compute-communication":[30],"overlap":[31],"but":[32],"often":[33],"fail":[34],"to":[35,146],"meet":[36],"theoretical":[37],"peak":[38],"performance":[39],"across":[40],"heterogeneous":[41],"new":[44],"accelerators.":[45],"Instead":[46],"of":[47,56,65,82,95,113,140],"operator-specific":[48],"techniques,":[49],"we":[50],"ask":[51],"whether":[52],"small":[54],"set":[55],"simple,":[57],"reusable":[58],"principles":[59,94],"can":[60],"systematically":[61],"guide":[62],"the":[63,80,88,93,114],"design":[64,98,124],"optimal":[66],"multi-GPU":[67,84,96,118],"kernels.":[68,85],"We":[69,126],"present":[70],"ParallelKittens":[71],"(PK),":[72],"minimal":[74],"CUDA":[75],"framework":[76,90],"that":[77,116],"drastically":[78],"simplifies":[79],"development":[81],"overlapped":[83],"PK":[86,128,143],"extends":[87],"ThunderKittens":[89],"embodies":[92],"kernel":[97],"eight":[100],"core":[101],"primitives":[102],"unified":[105],"programming":[106],"template,":[107],"derived":[108],"from":[109],"comprehensive":[111],"analysis":[112],"factors":[115],"govern":[117],"performance$\\unicode{x2014}$data-transfer":[119],"mechanisms,":[120],"resource":[121],"scheduling,":[122],"overheads.":[125],"validate":[127],"on":[129],"both":[130],"Hopper":[131],"Blackwell":[133],"architectures.":[134],"With":[135],"fewer":[136],"than":[137],"50":[138],"lines":[139],"device":[141],"code,":[142],"achieves":[144],"up":[145],"$2.33":[147],"\\times$":[148,156,162],"speedup":[149],"data-":[151],"tensor-parallel":[153],"workloads,":[154,159],"$4.08":[155],"sequence-parallel":[158],"$1.22":[161],"expert-parallel":[164],"workloads.":[165]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2025-11-20T00:00:00"}