{"id":"https://openalex.org/W4400977324","doi":"https://doi.org/10.1109/mm.2024.3428548","title":"Composition of Experts on the SN40L Reconfigurable Dataflow Unit","display_name":"Composition of Experts on the SN40L Reconfigurable Dataflow Unit","publication_year":2024,"publication_date":"2024-07-25","ids":{"openalex":"https://openalex.org/W4400977324","doi":"https://doi.org/10.1109/mm.2024.3428548"},"language":"en","primary_location":{"id":"doi:10.1109/mm.2024.3428548","is_oa":false,"landing_page_url":"https://doi.org/10.1109/mm.2024.3428548","pdf_url":null,"source":{"id":"https://openalex.org/S59697426","display_name":"IEEE Micro","issn_l":"0272-1732","issn":["0272-1732","1937-4143"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Micro","raw_type":"journal-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5111746097","display_name":"Raghu Prabhakar","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Raghu Prabhakar","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":"https://orcid.org/0000-0003-0230-4377","affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065493508","display_name":"Ram Sivaramakrishnan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ram Sivaramakrishnan","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Darshan Gandhi","orcid":"https://orcid.org/0009-0002-9442-4228"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Darshan Gandhi","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":"https://orcid.org/0009-0002-9442-4228","affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5109960019","display_name":"Yun Du","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yun Du","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5112914114","display_name":"Mingran Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mingran Wang","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Xiangyu Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiangyu Song","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101503011","display_name":"Kejie Zhang","orcid":"https://orcid.org/0000-0001-9783-4412"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kejie Zhang","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5089291480","display_name":"Tianren Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tianren Gao","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":"https://orcid.org/0009-0006-5144-5183","affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5048145208","display_name":"Angela Yee\u2010Moon Wang","orcid":"https://orcid.org/0000-0003-2508-7117"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Angela Wang","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5073949362","display_name":"Karen Li","orcid":"https://orcid.org/0000-0002-0261-743X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiaoyan Li","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5098316663","display_name":"Joshua Brot","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Joshua Brot","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Calvin Leung","orcid":"https://orcid.org/0009-0002-4134-3803"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Calvin Leung","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":"https://orcid.org/0009-0002-4134-3803","affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033875105","display_name":"Tuowen Zhao","orcid":"https://orcid.org/0000-0002-6347-7135"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tuowen Zhao","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5057532402","display_name":"Mark Gottscho","orcid":"https://orcid.org/0000-0001-8370-4158"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Mark Gottscho","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":null,"display_name":"Zhengyu Chen","orcid":"https://orcid.org/0000-0001-5811-456X"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhengyu Chen","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":"https://orcid.org/0000-0001-5811-456X","affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5062351534","display_name":"Kaizhao Liang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kaizhao Liang","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5034858834","display_name":"Swayambhoo Jain","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Swayambhoo Jain","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5090396089","display_name":"Urmish Thakker","orcid":"https://orcid.org/0000-0002-0515-9155"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Urmish Thakker","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103616447","display_name":"Kevin J. Brown","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kevin J. Brown","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5023857198","display_name":"Kunle Olukotun","orcid":"https://orcid.org/0000-0002-8779-0636"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kunle Olukotun","raw_affiliation_strings":["SambaNova Systems Inc, Palo Alto, CA, USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"SambaNova Systems Inc, Palo Alto, CA, USA","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":20,"corresponding_author_ids":["https://openalex.org/A5111746097"],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.16031469,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"44","issue":"6","first_page":"34","last_page":"43"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10703","display_name":"Business Process Modeling and Analysis","score":0.9553999900817871,"subfield":{"id":"https://openalex.org/subfields/1404","display_name":"Management Information Systems"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},"topics":[{"id":"https://openalex.org/T10703","display_name":"Business Process Modeling and Analysis","score":0.9553999900817871,"subfield":{"id":"https://openalex.org/subfields/1404","display_name":"Management Information Systems"},"field":{"id":"https://openalex.org/fields/14","display_name":"Business, Management and Accounting"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}},{"id":"https://openalex.org/T11986","display_name":"Scientific Computing and Data Management","score":0.9387999773025513,"subfield":{"id":"https://openalex.org/subfields/1802","display_name":"Information Systems and Management"},"field":{"id":"https://openalex.org/fields/18","display_name":"Decision Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8664752244949341},{"id":"https://openalex.org/keywords/dataflow","display_name":"Dataflow","score":0.8065304756164551},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.48701053857803345},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.3655768036842346}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8664752244949341},{"id":"https://openalex.org/C96324660","wikidata":"https://www.wikidata.org/wiki/Q205446","display_name":"Dataflow","level":2,"score":0.8065304756164551},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.48701053857803345},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.3655768036842346}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/mm.2024.3428548","is_oa":false,"landing_page_url":"https://doi.org/10.1109/mm.2024.3428548","pdf_url":null,"source":{"id":"https://openalex.org/S59697426","display_name":"IEEE Micro","issn_l":"0272-1732","issn":["0272-1732","1937-4143"],"is_oa":false,"is_in_doaj":false,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Micro","raw_type":"journal-article"}],"best_oa_location":null,"sustainable_development_goals":[{"display_name":"Affordable and clean energy","score":0.4399999976158142,"id":"https://metadata.un.org/sdg/7"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":5,"referenced_works":["https://openalex.org/W4213249026","https://openalex.org/W6847076894","https://openalex.org/W6852201467","https://openalex.org/W6858619786","https://openalex.org/W6866860863"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2293118914","https://openalex.org/W2998381397","https://openalex.org/W4236419692","https://openalex.org/W3167919718","https://openalex.org/W4251718783","https://openalex.org/W2171015181","https://openalex.org/W4239447582","https://openalex.org/W1998888015"],"abstract_inverted_index":{"Monolithic":[0],"large":[1],"language":[2],"models":[3],"(LLMs)":[4],"pose":[5],"significant":[6],"challenges":[7,43],"in":[8],"training":[9,33],"and":[10,30,34,52,72,85,100,133,144],"serving":[11],"during":[12],"an":[13,108,135],"active":[14],"deployment.":[15],"By":[16],"contrast,":[17],"composition":[18],"of":[19,32,55,138],"experts":[20],"(CoE)":[21],"is":[22],"a":[23,73,116,121,141,147],"modular":[24],"approach":[25],"that":[26],"lowers":[27],"the":[28,53,61],"cost":[29,54],"complexity":[31],"serving.":[35],"In":[36],"this":[37],"article,":[38],"we":[39],"explore":[40],"unique":[41],"hardware":[42],"for":[44],"CoE":[45,119],"models,":[46],"such":[47],"as":[48],"lower":[49],"operational":[50],"intensity":[51],"switching":[56,129],"between":[57,98],"models.":[58],"We":[59],"describe":[60],"SambaNova":[62],"SN40L":[63,94,112],"Reconfigurable":[64],"Dataflow":[65],"Unit":[66],"(RDU),":[67],"which":[68],"combines":[69],"streaming":[70],"dataflow":[71],"new":[74],"three-tier":[75],"memory":[76,81],"system":[77],"with":[78,120],"static":[79],"random-access":[80],"(RAM),":[82],"high-bandwidth":[83],"memory,":[84],"double":[86],"data":[87],"rate":[88],"dynamic":[89],"RAM.":[90],"A":[91],"single":[92],"eight-socket":[93],"node":[95,113],"achieves":[96,134],"speedups":[97],"2":[99],"13\u00d7":[101],"due":[102],"to":[103],"aggressive":[104],"operator":[105],"fusion":[106],"over":[107,140,146],"optimized":[109],"baseline.":[110],"The":[111],"deploys":[114],"Samba-CoE,":[115],"1":[117],"trillion-parameter":[118],"19\u00d7":[122],"smaller":[123],"machine":[124],"footprint,":[125],"speeds":[126],"up":[127],"model":[128],"time":[130],"by":[131],"15\u201331\u00d7":[132],"overall":[136],"speedup":[137],"3.7\u00d7":[139],"DGX":[142,148],"H100":[143],"6.6\u00d7":[145],"A100.":[149]},"counts_by_year":[],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}
