{"id":"https://openalex.org/W4416004084","doi":"https://doi.org/10.1145/3731599.3767497","title":"Roofline Analysis of Tightly-Coupled CPU-GPU Superchips: A Study on MI300A and GH200","display_name":"Roofline Analysis of Tightly-Coupled CPU-GPU Superchips: A Study on MI300A and GH200","publication_year":2025,"publication_date":"2025-11-07","ids":{"openalex":"https://openalex.org/W4416004084","doi":"https://doi.org/10.1145/3731599.3767497"},"language":null,"primary_location":{"id":"doi:10.1145/3731599.3767497","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767497","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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3731599.3767497","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5035719658","display_name":"Oscar Antepara","orcid":"https://orcid.org/0000-0002-4596-0289"},"institutions":[{"id":"https://openalex.org/I148283060","display_name":"Lawrence Berkeley National Laboratory","ror":"https://ror.org/02jbv0t02","country_code":"US","type":"facility","lineage":["https://openalex.org/I1330989302","https://openalex.org/I148283060","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Oscar Antepara","raw_affiliation_strings":["AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA"],"affiliations":[{"raw_affiliation_string":"AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA","institution_ids":["https://openalex.org/I148283060"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5113842004","display_name":"Leonid Oliker","orcid":"https://orcid.org/0000-0002-7923-2896"},"institutions":[{"id":"https://openalex.org/I148283060","display_name":"Lawrence Berkeley National Laboratory","ror":"https://ror.org/02jbv0t02","country_code":"US","type":"facility","lineage":["https://openalex.org/I1330989302","https://openalex.org/I148283060","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Leonid Oliker","raw_affiliation_strings":["AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA"],"affiliations":[{"raw_affiliation_string":"AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA","institution_ids":["https://openalex.org/I148283060"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5102746800","display_name":"Samuel Williams","orcid":"https://orcid.org/0000-0002-8327-5717"},"institutions":[{"id":"https://openalex.org/I148283060","display_name":"Lawrence Berkeley National Laboratory","ror":"https://ror.org/02jbv0t02","country_code":"US","type":"facility","lineage":["https://openalex.org/I1330989302","https://openalex.org/I148283060","https://openalex.org/I39565521"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Samuel Williams","raw_affiliation_strings":["AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA"],"affiliations":[{"raw_affiliation_string":"AMCR, Lawrence Berkeley National Laboratory, Berkeley, USA","institution_ids":["https://openalex.org/I148283060"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5035719658"],"corresponding_institution_ids":["https://openalex.org/I148283060"],"apc_list":null,"apc_paid":null,"fwci":2.5505,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.9089557,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":91,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"1205","last_page":"1216"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9156000018119812,"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":0.9156000018119812,"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/T11181","display_name":"Advanced Data Storage Technologies","score":0.029200000688433647,"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/T10101","display_name":"Cloud Computing and Resource Management","score":0.01590000092983246,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/software-portability","display_name":"Software portability","score":0.7828999757766724},{"id":"https://openalex.org/keywords/bottleneck","display_name":"Bottleneck","score":0.7411999702453613},{"id":"https://openalex.org/keywords/allocator","display_name":"Allocator","score":0.5439000129699707},{"id":"https://openalex.org/keywords/pci-express","display_name":"PCI Express","score":0.5436000227928162},{"id":"https://openalex.org/keywords/host","display_name":"Host (biology)","score":0.4018999934196472},{"id":"https://openalex.org/keywords/randomness","display_name":"Randomness","score":0.39100000262260437},{"id":"https://openalex.org/keywords/bandwidth","display_name":"Bandwidth (computing)","score":0.3873000144958496},{"id":"https://openalex.org/keywords/architecture","display_name":"Architecture","score":0.3756999969482422},{"id":"https://openalex.org/keywords/execution-model","display_name":"Execution model","score":0.3610999882221222}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8043000102043152},{"id":"https://openalex.org/C63000827","wikidata":"https://www.wikidata.org/wiki/Q3080428","display_name":"Software portability","level":2,"score":0.7828999757766724},{"id":"https://openalex.org/C2780513914","wikidata":"https://www.wikidata.org/wiki/Q18210350","display_name":"Bottleneck","level":2,"score":0.7411999702453613},{"id":"https://openalex.org/C162262903","wikidata":"https://www.wikidata.org/wiki/Q343527","display_name":"Allocator","level":2,"score":0.5439000129699707},{"id":"https://openalex.org/C64270927","wikidata":"https://www.wikidata.org/wiki/Q206924","display_name":"PCI Express","level":3,"score":0.5436000227928162},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4691999852657318},{"id":"https://openalex.org/C126831891","wikidata":"https://www.wikidata.org/wiki/Q221673","display_name":"Host (biology)","level":2,"score":0.4018999934196472},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.39800000190734863},{"id":"https://openalex.org/C125112378","wikidata":"https://www.wikidata.org/wiki/Q176640","display_name":"Randomness","level":2,"score":0.39100000262260437},{"id":"https://openalex.org/C2776257435","wikidata":"https://www.wikidata.org/wiki/Q1576430","display_name":"Bandwidth (computing)","level":2,"score":0.3873000144958496},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.38190001249313354},{"id":"https://openalex.org/C123657996","wikidata":"https://www.wikidata.org/wiki/Q12271","display_name":"Architecture","level":2,"score":0.3756999969482422},{"id":"https://openalex.org/C2776834041","wikidata":"https://www.wikidata.org/wiki/Q25346349","display_name":"Execution model","level":2,"score":0.3610999882221222},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.3440999984741211},{"id":"https://openalex.org/C31395832","wikidata":"https://www.wikidata.org/wiki/Q1318674","display_name":"Testbed","level":2,"score":0.34150001406669617},{"id":"https://openalex.org/C133875982","wikidata":"https://www.wikidata.org/wiki/Q764810","display_name":"Shared memory","level":2,"score":0.33180001378059387},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.3212999999523163},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.31279999017715454},{"id":"https://openalex.org/C188045654","wikidata":"https://www.wikidata.org/wiki/Q17148339","display_name":"Memory bandwidth","level":2,"score":0.30399999022483826},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.2946000099182129},{"id":"https://openalex.org/C28034677","wikidata":"https://www.wikidata.org/wiki/Q17092530","display_name":"Interleaving","level":2,"score":0.289000004529953},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.2791999876499176},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.26820001006126404},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.26739999651908875},{"id":"https://openalex.org/C202491316","wikidata":"https://www.wikidata.org/wiki/Q272683","display_name":"Instruction set","level":2,"score":0.26330000162124634},{"id":"https://openalex.org/C204156049","wikidata":"https://www.wikidata.org/wiki/Q751436","display_name":"Inter-process communication","level":2,"score":0.26179999113082886},{"id":"https://openalex.org/C2779602883","wikidata":"https://www.wikidata.org/wiki/Q15544750","display_name":"Memory architecture","level":2,"score":0.26030001044273376},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.25769999623298645},{"id":"https://openalex.org/C130795937","wikidata":"https://www.wikidata.org/wiki/Q2561570","display_name":"Remote direct memory access","level":2,"score":0.25290000438690186},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.25029999017715454}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3731599.3767497","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767497","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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3731599.3767497","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3731599.3767497","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 SC '25 Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":18,"referenced_works":["https://openalex.org/W842872497","https://openalex.org/W2002555321","https://openalex.org/W2607112060","https://openalex.org/W2743401659","https://openalex.org/W2983865192","https://openalex.org/W3034645333","https://openalex.org/W3046786790","https://openalex.org/W4360605382","https://openalex.org/W4385581820","https://openalex.org/W4396814805","https://openalex.org/W4401211711","https://openalex.org/W4401408748","https://openalex.org/W4401408779","https://openalex.org/W4405861433","https://openalex.org/W4406157083","https://openalex.org/W4409714147","https://openalex.org/W4412150154","https://openalex.org/W7105215290"],"related_works":[],"abstract_inverted_index":{"The":[0],"introduction":[1],"of":[2,46,140,159,171,173,192,200],"tightly-coupled":[3,207],"heterogeneous":[4],"architectures,":[5],"such":[6],"as":[7,31,175,177],"AMD\u2019s":[8],"MI300A":[9,48,148],"and":[10,59,91,109,149,179,195,203],"NVIDIA\u2019s":[11],"Grace-Hopper":[12],"(GH200),":[13],"address":[14],"a":[15,32,82,187],"long-standing":[16],"bottleneck":[17],"in":[18,35,81,142,198],"accelerated":[19],"computing,":[20],"namely":[21],"the":[22,26,42,47,57,86,131,137,147,157,169,190,196,211],"CPU-GPU":[23,36],"interface.":[24],"Whereas":[25,85],"GH200":[27,150],"can":[28],"be":[29],"seen":[30],"technological":[33],"leap":[34],"connectivity":[37],"greatly":[38],"exceeding":[39],"PCIe":[40],"cadence,":[41],"unified":[43],"memory":[44,73,161],"architecture":[45],"APU":[49],"enables":[50],"seamless":[51],"communication":[52],"through":[53],"coherent":[54],"caches.":[55],"When":[56],"CPU":[58],"GPU":[60,193],"execute":[61],"concurrently,":[62],"they":[63,76,112],"contend":[64,78],"not":[65],"only":[66],"for":[67,79,102],"finite":[68],"bandwidth":[69],"on":[70,146,189],"their":[71],"shared":[72,94],"interfaces,":[74],"but":[75],"also":[77],"power":[80],"power-constrained":[83],"environment.":[84],"serialized":[87],"producer-consumer,":[88,90],"overlapped/pipelined":[89],"multi-tenancy":[92],"(i.e.,":[93],"virtualized":[95],"hardware)":[96],"execution":[97,144],"models":[98],"offer":[99],"clear":[100,115],"avenues":[101],"exploiting":[103],"these":[104],"architectures":[105,194],"to":[106,135,210],"increase":[107],"HPC":[108,212],"ML/AI":[110],"throughput,":[111],"come":[113],"with":[114,118,186],"risks":[116],"associated":[117],"non-linear":[119],"performance":[120,138,165],"degradation":[121],"arising":[122],"from":[123],"this":[124,127,154],"contention.":[125],"In":[126],"paper,":[128],"we":[129],"extend":[130],"well-established":[132],"Roofline":[133],"model":[134],"capture":[136],"implications":[139],"contention":[141],"concurrent":[143],"modalities":[145],"architectures.":[151],"We":[152,184],"enhance":[153],"by":[155],"quantifying":[156],"portability":[158,166],"different":[160],"allocators":[162],"using":[163],"standard":[164],"metrics,":[167],"noting":[168],"impact":[170,197],"randomness":[172],"data":[174],"well":[176],"host":[178],"device":[180],"arithmetic":[181],"(compute)":[182],"intensity.":[183],"conclude":[185],"discussion":[188],"evolution":[191],"terms":[199],"performance,":[201],"portability,":[202],"programmability":[204],"that":[205],"emerging":[206],"GPUs":[208],"bring":[209],"landscape.":[213]},"counts_by_year":[{"year":2025,"cited_by_count":1}],"updated_date":"2026-03-07T16:01:11.037858","created_date":"2025-11-07T00:00:00"}