{"id":"https://openalex.org/W4388561287","doi":"https://doi.org/10.1145/3624062.3624257","title":"Benchmarking and In-depth Performance Study of Large Language Models on Habana Gaudi Processors","display_name":"Benchmarking and In-depth Performance Study of Large Language Models on Habana Gaudi Processors","publication_year":2023,"publication_date":"2023-11-10","ids":{"openalex":"https://openalex.org/W4388561287","doi":"https://doi.org/10.1145/3624062.3624257"},"language":"en","primary_location":{"id":"doi:10.1145/3624062.3624257","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3624062.3624257","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the SC '23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis","raw_type":"proceedings-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/A5100691056","display_name":"Chengming Zhang","orcid":"https://orcid.org/0000-0003-3008-9133"},"institutions":[{"id":"https://openalex.org/I592451","display_name":"Indiana University","ror":"https://ror.org/01kg8sb98","country_code":"US","type":"education","lineage":["https://openalex.org/I592451"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Chengming Zhang","raw_affiliation_strings":["Indiana University, United States of America"],"affiliations":[{"raw_affiliation_string":"Indiana University, United States of America","institution_ids":["https://openalex.org/I592451"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5004397575","display_name":"Baixi Sun","orcid":"https://orcid.org/0000-0001-9807-7978"},"institutions":[{"id":"https://openalex.org/I592451","display_name":"Indiana University","ror":"https://ror.org/01kg8sb98","country_code":"US","type":"education","lineage":["https://openalex.org/I592451"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Baixi Sun","raw_affiliation_strings":["Indiana University, United States of America"],"affiliations":[{"raw_affiliation_string":"Indiana University, United States of America","institution_ids":["https://openalex.org/I592451"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052001478","display_name":"Xiaodong Yu","orcid":"https://orcid.org/0000-0001-6244-1264"},"institutions":[{"id":"https://openalex.org/I108468826","display_name":"Stevens Institute of Technology","ror":"https://ror.org/02z43xh36","country_code":"US","type":"education","lineage":["https://openalex.org/I108468826"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Xiaodong Yu","raw_affiliation_strings":["Stevens Institute of Technology, United States of America"],"affiliations":[{"raw_affiliation_string":"Stevens Institute of Technology, United States of America","institution_ids":["https://openalex.org/I108468826"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5037768390","display_name":"Zhen Xie","orcid":"https://orcid.org/0000-0003-3516-2192"},"institutions":[{"id":"https://openalex.org/I123946342","display_name":"Binghamton University","ror":"https://ror.org/008rmbt77","country_code":"US","type":"education","lineage":["https://openalex.org/I123946342"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhen Xie","raw_affiliation_strings":["Binghamton University, United States of America"],"affiliations":[{"raw_affiliation_string":"Binghamton University, United States of America","institution_ids":["https://openalex.org/I123946342"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5051499975","display_name":"Weijian Zheng","orcid":"https://orcid.org/0000-0003-2791-0031"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Weijian Zheng","raw_affiliation_strings":["Argonne National Laboratory, United States of America"],"affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, United States of America","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055862557","display_name":"Kamil Iskra","orcid":"https://orcid.org/0000-0001-7000-4195"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kamil A. Iskra","raw_affiliation_strings":["Argonne National Laboratory, United States of America"],"affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, United States of America","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5038177319","display_name":"Pete Beckman","orcid":"https://orcid.org/0000-0002-9428-7801"},"institutions":[{"id":"https://openalex.org/I1282105669","display_name":"Argonne National Laboratory","ror":"https://ror.org/05gvnxz63","country_code":"US","type":"facility","lineage":["https://openalex.org/I1282105669","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Pete Beckman","raw_affiliation_strings":["Argonne National Laboratory, United States of America"],"affiliations":[{"raw_affiliation_string":"Argonne National Laboratory, United States of America","institution_ids":["https://openalex.org/I1282105669"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5063703614","display_name":"Dingwen Tao","orcid":"https://orcid.org/0000-0001-5422-4497"},"institutions":[{"id":"https://openalex.org/I592451","display_name":"Indiana University","ror":"https://ror.org/01kg8sb98","country_code":"US","type":"education","lineage":["https://openalex.org/I592451"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dingwen Tao","raw_affiliation_strings":["Indiana University, United States of America"],"affiliations":[{"raw_affiliation_string":"Indiana University, United States of America","institution_ids":["https://openalex.org/I592451"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":8,"corresponding_author_ids":["https://openalex.org/A5100691056"],"corresponding_institution_ids":["https://openalex.org/I592451"],"apc_list":null,"apc_paid":null,"fwci":0.8741,"has_fulltext":false,"cited_by_count":5,"citation_normalized_percentile":{"value":0.79339826,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":98},"biblio":{"volume":null,"issue":null,"first_page":"1759","last_page":"1766"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.9879000186920166,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10028","display_name":"Topic Modeling","score":0.9879000186920166,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.9625999927520752,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"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/benchmarking","display_name":"Benchmarking","score":0.9116290807723999},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6744735836982727},{"id":"https://openalex.org/keywords/business","display_name":"Business","score":0.1396576464176178},{"id":"https://openalex.org/keywords/marketing","display_name":"Marketing","score":0.05279877781867981}],"concepts":[{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.9116290807723999},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6744735836982727},{"id":"https://openalex.org/C144133560","wikidata":"https://www.wikidata.org/wiki/Q4830453","display_name":"Business","level":0,"score":0.1396576464176178},{"id":"https://openalex.org/C162853370","wikidata":"https://www.wikidata.org/wiki/Q39809","display_name":"Marketing","level":1,"score":0.05279877781867981}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3624062.3624257","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3624062.3624257","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the SC '23 Workshops of the International Conference on High Performance Computing, Network, Storage, and Analysis","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"score":0.41999998688697815,"id":"https://metadata.un.org/sdg/9","display_name":"Industry, innovation and infrastructure"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":7,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W1566289585","https://openalex.org/W2963122961","https://openalex.org/W2979826702","https://openalex.org/W3006732000","https://openalex.org/W4283704460","https://openalex.org/W4321636820"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W4238897586","https://openalex.org/W435179959","https://openalex.org/W2619091065","https://openalex.org/W2059640416","https://openalex.org/W1490753184","https://openalex.org/W2284465472","https://openalex.org/W2291782699","https://openalex.org/W1993948687"],"abstract_inverted_index":{"Transformer":[0],"models":[1,159],"have":[2],"achieved":[3],"remarkable":[4],"success":[5],"in":[6,91,140],"various":[7],"machine":[8],"learning":[9],"tasks":[10],"but":[11],"suffer":[12],"from":[13],"high":[14],"computational":[15,128],"complexity":[16,22],"and":[17,36,64,104,111,120,144,173,187,196],"resource":[18],"requirements.":[19],"The":[20,163],"quadratic":[21],"of":[23,67,80,135,154,165],"the":[24,45,77,92,102,133,151,206],"self-attention":[25],"mechanism":[26],"further":[27],"exacerbates":[28],"these":[29,55],"challenges":[30,90],"when":[31],"dealing":[32],"with":[33],"long":[34,142],"sequences":[35,143],"large":[37,157],"datasets.":[38],"Specialized":[39],"AI":[40],"hardware":[41],"accelerators,":[42],"such":[43],"as":[44],"Habana":[46],"GAUDI":[47,57,82,207],"architecture,":[48],"offer":[49],"a":[50,59,65,97,193,198],"promising":[51],"solution":[52],"to":[53,84,117,126],"tackle":[54],"issues.":[56],"features":[58],"Matrix":[60],"Multiplication":[61],"Engine":[62],"(MME)":[63],"cluster":[66],"fully":[68],"programmable":[69],"Tensor":[70],"Processing":[71],"Cores":[72],"(TPC).":[73],"This":[74],"paper":[75],"explores":[76],"untapped":[78],"potential":[79],"using":[81],"processors":[83],"accelerate":[85],"Transformer-based":[86,156,202],"models,":[87],"addressing":[88],"key":[89],"process.":[93],"Firstly,":[94],"we":[95,114,131,149],"provide":[96],"comprehensive":[98],"performance":[99,134,146,153],"comparison":[100],"between":[101],"MME":[103,119],"TPC":[105,121],"components,":[106],"illuminating":[107],"their":[108],"relative":[109],"strengths":[110],"weaknesses.":[112],"Secondly,":[113],"explore":[115],"strategies":[116],"optimize":[118],"utilization,":[122],"offering":[123],"practical":[124,169],"insights":[125,170],"enhance":[127],"efficiency.":[129],"Thirdly,":[130],"evaluate":[132,150],"Transformers":[136,182],"on":[137,161,205],"GAUDI,":[138],"particularly":[139],"handling":[141],"uncovering":[145],"bottlenecks.":[147],"Lastly,":[148],"end-to-end":[152],"two":[155],"language":[158],"(LLM)":[160],"GAUDI.":[162],"contributions":[164],"this":[166],"work":[167],"encompass":[168],"for":[171,181,200],"practitioners":[172],"researchers":[174],"alike.":[175],"We":[176],"delve":[177],"into":[178],"GAUDI\u2019s":[179],"capabilities":[180],"through":[183],"systematic":[184],"profiling,":[185],"analysis,":[186],"optimization":[188],"exploration.":[189],"Our":[190],"study":[191],"bridges":[192],"research":[194],"gap":[195],"offers":[197],"roadmap":[199],"optimizing":[201],"model":[203],"training":[204],"architecture.":[208]},"counts_by_year":[{"year":2025,"cited_by_count":4},{"year":2024,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}