{"id":"https://openalex.org/W4416004242","doi":"https://doi.org/10.1145/3731599.3767391","title":"MPI Collectives with Programmable Smart Switches","display_name":"MPI Collectives with Programmable Smart Switches","publication_year":2025,"publication_date":"2025-11-07","ids":{"openalex":"https://openalex.org/W4416004242","doi":"https://doi.org/10.1145/3731599.3767391"},"language":"en","primary_location":{"id":"doi:10.1145/3731599.3767391","doi":"https://doi.org/10.1145/3731599.3767391","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767391","pdf_url":null,"source":{"id":"https://openalex.org/S4363608557","display_name":"International Conference for High Performance Computing, Networking, Storage and Analysis (SC)","issn_l":"2167-4337","issn":["2167-4337"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","is_traditional":true,"indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"hybrid","oa_url":"https://doi.org/10.1145/3731599.3767391","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5120303655","display_name":"Thomas Erbesdobler","orcid":"https://orcid.org/0009-0000-5310-4346","observed_orcids":["https://orcid.org/0009-0000-5310-4346"]},"institutions":[{"id":"https://openalex.org/I62916508","display_name":"Technical University of Munich","ror":"https://ror.org/02kkvpp62","country_code":"DE","type":"education","lineage":["https://openalex.org/I62916508"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Thomas Erbesdobler","raw_affiliation_strings":["Technical University of Munich, Munich, Germany"],"raw_orcid":"https://orcid.org/0009-0000-5310-4346","affiliations":[{"raw_affiliation_string":"Technical University of Munich, Munich, Germany","institution_ids":["https://openalex.org/I62916508"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5005482380","display_name":"Amir Raoofy","orcid":"https://orcid.org/0000-0001-9664-8298","observed_orcids":["https://orcid.org/0000-0001-9664-8298"]},"institutions":[{"id":"https://openalex.org/I4210163716","display_name":"Leibniz Supercomputing Centre","ror":"https://ror.org/05558nw16","country_code":"DE","type":"facility","lineage":["https://openalex.org/I109144446","https://openalex.org/I4210163716"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Amir Raoofy","raw_affiliation_strings":["Leibniz Supercomputing Centre (LRZ), Garching, Germany"],"raw_orcid":"https://orcid.org/0000-0001-9664-8298","affiliations":[{"raw_affiliation_string":"Leibniz Supercomputing Centre (LRZ), Garching, Germany","institution_ids":["https://openalex.org/I4210163716"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004451196","display_name":"Ehab Saleh","orcid":"https://orcid.org/0000-0002-4950-4854","observed_orcids":["https://orcid.org/0000-0002-4950-4854"]},"institutions":[{"id":"https://openalex.org/I4210163716","display_name":"Leibniz Supercomputing Centre","ror":"https://ror.org/05558nw16","country_code":"DE","type":"facility","lineage":["https://openalex.org/I109144446","https://openalex.org/I4210163716"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Ehab Saleh","raw_affiliation_strings":["Leibniz Supercomputing Centre (LRZ), Garching, Germany"],"raw_orcid":"https://orcid.org/0000-0002-4950-4854","affiliations":[{"raw_affiliation_string":"Leibniz Supercomputing Centre (LRZ), Garching, Germany","institution_ids":["https://openalex.org/I4210163716"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5073800722","display_name":"Josef Weidendorfer","orcid":"https://orcid.org/0000-0001-7159-1432","observed_orcids":["https://orcid.org/0000-0001-7159-1432"]},"institutions":[{"id":"https://openalex.org/I4210163716","display_name":"Leibniz Supercomputing Centre","ror":"https://ror.org/05558nw16","country_code":"DE","type":"facility","lineage":["https://openalex.org/I109144446","https://openalex.org/I4210163716"]}],"countries":["DE"],"is_corresponding":false,"raw_author_name":"Josef Weidendorfer","raw_affiliation_strings":["Leibniz Supercomputing Centre (LRZ), Garching b. M., Germany"],"raw_orcid":"https://orcid.org/0000-0001-7159-1432","affiliations":[{"raw_affiliation_string":"Leibniz Supercomputing Centre (LRZ), Garching b. M., Germany","institution_ids":["https://openalex.org/I4210163716"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.4748,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.67476418,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"468","last_page":"478"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.5952000021934509,"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":0.5952000021934509,"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/T10829","display_name":"Interconnection Networks and Systems","score":0.12370000034570694,"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"}},{"id":"https://openalex.org/T10772","display_name":"Distributed systems and fault tolerance","score":0.10779999941587448,"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/p4","display_name":"P4","score":0.7919999957084656},{"id":"https://openalex.org/keywords/offloading","display_name":"offloading","score":0.675000011920929},{"id":"https://openalex.org/keywords/reduction","display_name":"reduction","score":0.8500000238418579},{"id":"https://openalex.org/keywords/rocev2","display_name":"RoCEv2","score":0.8690000176429749},{"id":"https://openalex.org/keywords/rdma","display_name":"RDMA","score":0.9990000128746033},{"id":"https://openalex.org/keywords/open-mpi","display_name":"Open MPI","score":0.8820000290870667},{"id":"https://openalex.org/keywords/high-performance-computing","display_name":"high-performance computing","score":0.5920000076293945}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8022000193595886},{"id":"https://openalex.org/C130795937","wikidata":"https://www.wikidata.org/wiki/Q2561570","display_name":"Remote direct memory access","level":2,"score":0.8012999892234802},{"id":"https://openalex.org/C172173386","wikidata":"https://www.wikidata.org/wiki/Q79984","display_name":"Ethernet","level":2,"score":0.6187000274658203},{"id":"https://openalex.org/C79974875","wikidata":"https://www.wikidata.org/wiki/Q483639","display_name":"Cloud computing","level":2,"score":0.5101000070571899},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.4844000041484833},{"id":"https://openalex.org/C119404949","wikidata":"https://www.wikidata.org/wiki/Q4503","display_name":"Network switch","level":2,"score":0.47760000824928284},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4537999927997589},{"id":"https://openalex.org/C158379750","wikidata":"https://www.wikidata.org/wiki/Q214111","display_name":"Network packet","level":2,"score":0.4526999890804291},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.4413999915122986},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.4262000024318695},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.400299996137619},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.3955000042915344},{"id":"https://openalex.org/C159631557","wikidata":"https://www.wikidata.org/wiki/Q1546066","display_name":"Networking hardware","level":2,"score":0.3463999927043915},{"id":"https://openalex.org/C42935608","wikidata":"https://www.wikidata.org/wiki/Q190411","display_name":"Field-programmable gate array","level":2,"score":0.33340001106262207},{"id":"https://openalex.org/C51332947","wikidata":"https://www.wikidata.org/wiki/Q1172305","display_name":"Shared resource","level":2,"score":0.31349998712539673},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.30640000104904175},{"id":"https://openalex.org/C74366991","wikidata":"https://www.wikidata.org/wiki/Q2755335","display_name":"Network processor","level":3,"score":0.28949999809265137},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.28619998693466187},{"id":"https://openalex.org/C34165917","wikidata":"https://www.wikidata.org/wiki/Q188267","display_name":"Programming paradigm","level":2,"score":0.26980000734329224},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.2687999904155731},{"id":"https://openalex.org/C198352243","wikidata":"https://www.wikidata.org/wiki/Q37105","display_name":"Line (geometry)","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C138236772","wikidata":"https://www.wikidata.org/wiki/Q25098575","display_name":"Edge device","level":3,"score":0.2515000104904175}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3731599.3767391","doi":"https://doi.org/10.1145/3731599.3767391","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767391","pdf_url":null,"source":{"id":"https://openalex.org/S4363608557","display_name":"International Conference for High Performance Computing, Networking, Storage and Analysis (SC)","issn_l":"2167-4337","issn":["2167-4337"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3731599.3767391","doi":"https://doi.org/10.1145/3731599.3767391","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767391","pdf_url":null,"source":{"id":"https://openalex.org/S4363608557","display_name":"International Conference for High Performance Computing, Networking, Storage and Analysis (SC)","issn_l":"2167-4337","issn":["2167-4337"],"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":["jufo-1"],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure","score":0.428600013256073}],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W1742033587","https://openalex.org/W2143295630","https://openalex.org/W2479838429","https://openalex.org/W2550753940","https://openalex.org/W3012220622","https://openalex.org/W3089731223","https://openalex.org/W4242634435","https://openalex.org/W4296704754","https://openalex.org/W4396814849","https://openalex.org/W4402402782"],"related_works":["https://openalex.org/W7166310008","https://openalex.org/W3034261992","https://openalex.org/W123010799","https://openalex.org/W3113824144","https://openalex.org/W4379089615","https://openalex.org/W3160003613","https://openalex.org/W4399283082","https://openalex.org/W4200623295","https://openalex.org/W2761686942","https://openalex.org/W7161118312"],"abstract_inverted_index":{"Programmable":[0],"smart":[1,58],"network":[2],"devices":[3],"are":[4,28],"heavily":[5],"used":[6],"by":[7],"cloud":[8],"providers,":[9],"but":[10],"typically":[11],"not":[12],"for":[13,19,24,30],"HPC.":[14],"However,":[15],"they":[16],"provide":[17],"opportunities":[18],"off-loading":[20],"computations,":[21],"in":[22,34,122],"particular":[23,55],"collective":[25,84,120],"operations,":[26],"which":[27],"important":[29],"data":[31,90,138],"intensive":[32],"workloads":[33],"classic":[35],"HPC":[36],"and":[37,75,125],"ML":[38],"training.":[39],"In":[40],"this":[41],"paper,":[42],"we":[43,76,115,132],"present":[44],"a":[45,72,119],"prototype":[46],"called":[47],"mpitofino":[48,117,135],"to":[49,82,101,141],"enable":[50],"offloading":[51],"MPI":[52,124],"collectives":[53],"(in":[54],"reductions)":[56],"onto":[57],"switches":[59,69],"over":[60],"an":[61],"Ethernet":[62,68],"fabric.":[63],"We":[64,92],"target":[65],"Intel\u2019s":[66],"programmable":[67],"equipped":[70],"with":[71],"Tofino":[73],"ASIC,":[74],"use":[77,102],"the":[78,87,95,111,142],"P4":[79,98],"programming":[80],"language":[81],"process":[83],"packets":[85],"on":[86,110],"chip\u2019s":[88],"low-latency":[89],"path.":[91],"demonstrate":[93,133],"how":[94],"flexibility":[96],"of":[97],"enables":[99],"us":[100],"RoCEv2":[103],"as":[104,118],"protocol,":[105],"utilizing":[106],"RDMA":[107],"hardware":[108],"support":[109],"nodes\u2019":[112],"NICs.":[113],"Furthermore,":[114],"implement":[116],"provider":[121],"Open":[123],"discuss":[126],"its":[127],"desirable":[128],"scaling":[129],"characteristics.":[130],"Finally,":[131],"that":[134],"can":[136],"achieve":[137],"throughput":[139],"close":[140],"100GBit/s":[143],"line":[144],"rate.":[145]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2025-11-07T00:00:00"}