{"id":"https://openalex.org/W7168029408","doi":"https://doi.org/10.1145/3806645.3816737","title":"HAT-MPI: Hierarchical Auto Tuning of MPI Inter-Node Communication on InfiniBand Clusters","display_name":"HAT-MPI: Hierarchical Auto Tuning of MPI Inter-Node Communication on InfiniBand Clusters","publication_year":2026,"publication_date":"2026-07-11","ids":{"openalex":"https://openalex.org/W7168029408","doi":"https://doi.org/10.1145/3806645.3816737"},"language":null,"primary_location":{"id":"doi:10.1145/3806645.3816737","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3806645.3816737","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 35th International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3806645.3816737","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5001165357","display_name":"Sungjae Lee","orcid":"https://orcid.org/0000-0002-8824-2164"},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sungjae Lee","raw_affiliation_strings":["The Ohio State University, Columbus, USA"],"raw_orcid":"https://orcid.org/0000-0002-8824-2164","affiliations":[{"raw_affiliation_string":"The Ohio State University, Columbus, USA","institution_ids":["https://openalex.org/I52357470"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5140465556","display_name":"Sam Tilford","orcid":"https://orcid.org/0009-0007-7849-6674"},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sam Tilford","raw_affiliation_strings":["The Ohio State University, Columbus, USA"],"raw_orcid":"https://orcid.org/0009-0007-7849-6674","affiliations":[{"raw_affiliation_string":"The Ohio State University, Columbus, USA","institution_ids":["https://openalex.org/I52357470"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5024879682","display_name":"Dhabaleswar K. Panda","orcid":"https://orcid.org/0000-0002-0356-1781"},"institutions":[{"id":"https://openalex.org/I52357470","display_name":"The Ohio State University","ror":"https://ror.org/00rs6vg23","country_code":"US","type":"education","lineage":["https://openalex.org/I52357470"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dhabaleswar K. Panda","raw_affiliation_strings":["The Ohio State University, Columbus, USA"],"raw_orcid":"https://orcid.org/0000-0002-0356-1781","affiliations":[{"raw_affiliation_string":"The Ohio State University, Columbus, USA","institution_ids":["https://openalex.org/I52357470"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I52357470"],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.93589165,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"731","last_page":"736"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.791100025177002,"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":0.791100025177002,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.05290000140666962,"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/T10715","display_name":"Distributed and Parallel Computing Systems","score":0.027899999171495438,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/infiniband","display_name":"InfiniBand","score":0.9922999739646912},{"id":"https://openalex.org/keywords/remote-direct-memory-access","display_name":"Remote direct memory access","score":0.6110000014305115},{"id":"https://openalex.org/keywords/protocol","display_name":"Protocol (science)","score":0.5871999859809875},{"id":"https://openalex.org/keywords/middleware","display_name":"Middleware (distributed applications)","score":0.5440999865531921},{"id":"https://openalex.org/keywords/interface","display_name":"Interface (matter)","score":0.5145000219345093},{"id":"https://openalex.org/keywords/selection","display_name":"Selection (genetic algorithm)","score":0.4997999966144562},{"id":"https://openalex.org/keywords/message-passing-interface","display_name":"Message Passing Interface","score":0.49050000309944153},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.4772999882698059}],"concepts":[{"id":"https://openalex.org/C2781030343","wikidata":"https://www.wikidata.org/wiki/Q922437","display_name":"InfiniBand","level":2,"score":0.9922999739646912},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8486999869346619},{"id":"https://openalex.org/C130795937","wikidata":"https://www.wikidata.org/wiki/Q2561570","display_name":"Remote direct memory access","level":2,"score":0.6110000014305115},{"id":"https://openalex.org/C2780385302","wikidata":"https://www.wikidata.org/wiki/Q367158","display_name":"Protocol (science)","level":3,"score":0.5871999859809875},{"id":"https://openalex.org/C169468491","wikidata":"https://www.wikidata.org/wiki/Q146923","display_name":"Middleware (distributed applications)","level":2,"score":0.5440999865531921},{"id":"https://openalex.org/C113843644","wikidata":"https://www.wikidata.org/wiki/Q901882","display_name":"Interface (matter)","level":4,"score":0.5145000219345093},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4997999966144562},{"id":"https://openalex.org/C166782233","wikidata":"https://www.wikidata.org/wiki/Q127879","display_name":"Message Passing Interface","level":3,"score":0.49050000309944153},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.48080000281333923},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.4772999882698059},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.43939998745918274},{"id":"https://openalex.org/C12269588","wikidata":"https://www.wikidata.org/wiki/Q132364","display_name":"Communications protocol","level":2,"score":0.4092000126838684},{"id":"https://openalex.org/C854659","wikidata":"https://www.wikidata.org/wiki/Q1859284","display_name":"Message passing","level":2,"score":0.40209999680519104},{"id":"https://openalex.org/C2775973920","wikidata":"https://www.wikidata.org/wiki/Q3252726","display_name":"Selection algorithm","level":3,"score":0.35690000653266907},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.31940001249313354},{"id":"https://openalex.org/C2986800882","wikidata":"https://www.wikidata.org/wiki/Q7168187","display_name":"Performance enhancement","level":2,"score":0.3156000077724457},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3089999854564667},{"id":"https://openalex.org/C133462117","wikidata":"https://www.wikidata.org/wiki/Q4929239","display_name":"Data collection","level":2,"score":0.2897999882698059},{"id":"https://openalex.org/C557945733","wikidata":"https://www.wikidata.org/wiki/Q389772","display_name":"Data transmission","level":2,"score":0.2800000011920929},{"id":"https://openalex.org/C103987645","wikidata":"https://www.wikidata.org/wiki/Q985806","display_name":"Network interface","level":3,"score":0.26910001039505005},{"id":"https://openalex.org/C73555534","wikidata":"https://www.wikidata.org/wiki/Q622825","display_name":"Cluster analysis","level":2,"score":0.2687000036239624},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.26409998536109924},{"id":"https://openalex.org/C123640240","wikidata":"https://www.wikidata.org/wiki/Q1092177","display_name":"Message oriented middleware","level":4,"score":0.26089999079704285},{"id":"https://openalex.org/C101765175","wikidata":"https://www.wikidata.org/wiki/Q577764","display_name":"Communications system","level":2,"score":0.2531999945640564},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.2517000138759613},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.25130000710487366}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3806645.3816737","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3806645.3816737","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 35th International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3806645.3816737","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3806645.3816737","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 35th International Symposium on High-Performance Parallel and Distributed Computing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G3358531649","display_name":"Collaborative Research: Frameworks: Performance Engineering Scientific Applications with MVAPICH and TAU using Emerging Communication Primitives","funder_award_id":"2311830","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G5950140645","display_name":"OAC Core: AI4MPI: ML-Based Optimization for MPI Library","funder_award_id":"2504944","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G8189731814","display_name":"Construction for the Leadership Class Computing Facility","funder_award_id":"2323116","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G8438779844","display_name":"CSR: Small: CONCERT: Designing Scalable Communication Runtimes with On-the-fly Compression for HPC and AI Applications on Heterogeneous Architectures","funder_award_id":"2312927","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"},{"id":"https://openalex.org/G855500582","display_name":"Collaborative Research: NSF-MeitY: CSR: Small: Eco-LLM: Energy-Efficient Computation and Communication for Large Language Models with  CXL-based Chip Architecture and Software","funder_award_id":"2415201","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":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":16,"referenced_works":["https://openalex.org/W1572549641","https://openalex.org/W1825216778","https://openalex.org/W1929619420","https://openalex.org/W1962931680","https://openalex.org/W1969923711","https://openalex.org/W1996020380","https://openalex.org/W2108244650","https://openalex.org/W3083442288","https://openalex.org/W3097072756","https://openalex.org/W4312268627","https://openalex.org/W4318603806","https://openalex.org/W4388581193","https://openalex.org/W4401017780","https://openalex.org/W4401017920","https://openalex.org/W4406157733","https://openalex.org/W7139920521"],"related_works":[],"abstract_inverted_index":{"Optimizing":[0],"point-to-point":[1],"and":[2,30,55,76,88,103,126,134,164],"collective":[3,79],"communication":[4,26,58],"in":[5,155],"HPC":[6,47,83,131],"systems":[7],"substantially":[8],"impacts":[9],"the":[10,43,71,77,99,104,122,143,162,168],"performance":[11,53,89,153],"of":[12,45,151],"large-scale":[13],"applications.":[14],"Middleware":[15],"libraries":[16],"such":[17],"as":[18],"Message":[19],"Passing":[20],"Interface":[21],"(MPI)":[22],"rely":[23],"on":[24,42,81],"lower-level":[25],"libraries,":[27],"including":[28],"UCX":[29,72,100,145],"OFI":[31],"Libfabric,":[32],"to":[33],"enable":[34],"these":[35],"optimizations.":[36],"However,":[37],"because":[38],"optimal":[39,78,105],"configurations":[40],"depend":[41],"characteristics":[44,87],"each":[46],"system,":[48],"users":[49],"must":[50],"understand":[51],"system-specific":[52],"behavior":[54],"manually":[56],"tune":[57],"parameters.":[59],"In":[60],"this":[61],"work,":[62],"we":[63],"present":[64],"HAT-MPI,":[65],"an":[66,149],"ML-based":[67],"framework":[68],"that":[69,120,142,178],"identifies":[70],"Eager-Rendezvous":[73],"protocol":[74,101,146],"threshold":[75,102,147],"algorithm":[80,107,137,169],"InfiniBand":[82,111],"clusters.":[84,112],"Using":[85],"hardware":[86,132],"data":[90,124],"collected":[91],"across":[92],"clusters":[93],"at":[94],"different":[95],"scales,":[96],"HAT-MPI":[97,179],"predicts":[98],"Alltoall/Allreduce":[106],"for":[108],"previously":[109],"unseen":[110],"This":[113],"paper":[114],"also":[115],"introduces":[116],"two":[117],"complementary":[118],"techniques":[119],"improve":[121],"conventional":[123],"collection":[125],"evaluation":[127],"workflow,":[128],"namely":[129],"LLM-assisted":[130],"detection":[133],"Top-5%":[135,175],"candidate":[136],"evaluation.":[138],"Experimental":[139],"results":[140],"show":[141],"predicted":[144,163],"achieves":[148],"average":[150],"31%":[152],"improvement":[154],"regions":[156],"where":[157],"a":[158],"gap":[159],"exists":[160],"between":[161],"default":[165],"thresholds,":[166],"while":[167],"selection":[170],"attains":[171],"84.9%":[172],"accuracy":[173],"with":[174],"evaluation,":[176],"demonstrating":[177],"delivers":[180],"highly":[181],"accurate":[182],"predictions.":[183]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-12T00:00:00"}