{"id":"https://openalex.org/W4416043136","doi":"https://doi.org/10.1109/ipdps65963.2026.00078","title":"Characterizing Production GPU Workloads using System-wide Telemetry Data","display_name":"Characterizing Production GPU Workloads using System-wide Telemetry Data","publication_year":2026,"publication_date":"2026-05-25","ids":{"openalex":"https://openalex.org/W4416043136","doi":"https://doi.org/10.1109/ipdps65963.2026.00078"},"language":"en","primary_location":{"id":"doi:10.1109/ipdps65963.2026.00078","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdps65963.2026.00078","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Parallel and Distributed Processing Symposium (IPDPS)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2502.18680","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5024124489","display_name":"Onur Cankur","orcid":null},"institutions":[{"id":"https://openalex.org/I66946132","display_name":"University of Maryland, College Park","ror":"https://ror.org/047s2c258","country_code":"US","type":"education","lineage":["https://openalex.org/I66946132"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Onur Cankur","raw_affiliation_strings":["University of Maryland,Department of Computer Science,College Park,Maryland,USA,20742"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Maryland,Department of Computer Science,College Park,Maryland,USA,20742","institution_ids":["https://openalex.org/I66946132"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5055945855","display_name":"Brian Austin","orcid":"https://orcid.org/0009-0005-5881-1927"},"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"]},{"id":"https://openalex.org/I4210151627","display_name":"National Energy Research Scientific Computing Center","ror":"https://ror.org/05v3mvq14","country_code":"US","type":"facility","lineage":["https://openalex.org/I1330989302","https://openalex.org/I148283060","https://openalex.org/I39565521","https://openalex.org/I4210151627"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Brian Austin","raw_affiliation_strings":["Lawrence Berkeley National Laboratory,NERSC,Berkeley,California,USA,94720"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Lawrence Berkeley National Laboratory,NERSC,Berkeley,California,USA,94720","institution_ids":["https://openalex.org/I148283060","https://openalex.org/I4210151627"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Dhruva Kulkarni","orcid":null},"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"]},{"id":"https://openalex.org/I4210151627","display_name":"National Energy Research Scientific Computing Center","ror":"https://ror.org/05v3mvq14","country_code":"US","type":"facility","lineage":["https://openalex.org/I1330989302","https://openalex.org/I148283060","https://openalex.org/I39565521","https://openalex.org/I4210151627"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Dhruva Kulkarni","raw_affiliation_strings":["Lawrence Berkeley National Laboratory,NERSC,Berkeley,California,USA,94720"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Lawrence Berkeley National Laboratory,NERSC,Berkeley,California,USA,94720","institution_ids":["https://openalex.org/I148283060","https://openalex.org/I4210151627"]}]},{"author_position":"last","author":{"id":null,"display_name":"Abhinav Bhatele","orcid":null},"institutions":[{"id":"https://openalex.org/I66946132","display_name":"University of Maryland, College Park","ror":"https://ror.org/047s2c258","country_code":"US","type":"education","lineage":["https://openalex.org/I66946132"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Abhinav Bhatele","raw_affiliation_strings":["University of Maryland,Department of Computer Science,College Park,Maryland,USA,20742"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Maryland,Department of Computer Science,College Park,Maryland,USA,20742","institution_ids":["https://openalex.org/I66946132"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"899","last_page":"912"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.7562999725341797,"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":0.7562999725341797,"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":0.15690000355243683,"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":0.02410000003874302,"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/workload","display_name":"Workload","score":0.8223000168800354},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.5605000257492065},{"id":"https://openalex.org/keywords/range","display_name":"Range (aeronautics)","score":0.46860000491142273},{"id":"https://openalex.org/keywords/production","display_name":"Production (economics)","score":0.4674000144004822},{"id":"https://openalex.org/keywords/service","display_name":"Service (business)","score":0.4641000032424927},{"id":"https://openalex.org/keywords/work","display_name":"Work (physics)","score":0.4632999897003174},{"id":"https://openalex.org/keywords/resource","display_name":"Resource (disambiguation)","score":0.4011000096797943},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.37470000982284546},{"id":"https://openalex.org/keywords/resource-allocation","display_name":"Resource allocation","score":0.3337000012397766}],"concepts":[{"id":"https://openalex.org/C2778476105","wikidata":"https://www.wikidata.org/wiki/Q628539","display_name":"Workload","level":2,"score":0.8223000168800354},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7961999773979187},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.5605000257492065},{"id":"https://openalex.org/C204323151","wikidata":"https://www.wikidata.org/wiki/Q905424","display_name":"Range (aeronautics)","level":2,"score":0.46860000491142273},{"id":"https://openalex.org/C2778348673","wikidata":"https://www.wikidata.org/wiki/Q739302","display_name":"Production (economics)","level":2,"score":0.4674000144004822},{"id":"https://openalex.org/C2780378061","wikidata":"https://www.wikidata.org/wiki/Q25351891","display_name":"Service (business)","level":2,"score":0.4641000032424927},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.4632999897003174},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.45210000872612},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.4016999900341034},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.4011000096797943},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.37470000982284546},{"id":"https://openalex.org/C29202148","wikidata":"https://www.wikidata.org/wiki/Q287260","display_name":"Resource allocation","level":2,"score":0.3337000012397766},{"id":"https://openalex.org/C186370098","wikidata":"https://www.wikidata.org/wiki/Q442787","display_name":"Energy (signal processing)","level":2,"score":0.3294999897480011},{"id":"https://openalex.org/C2742236","wikidata":"https://www.wikidata.org/wiki/Q924713","display_name":"Efficient energy use","level":2,"score":0.31189998984336853},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.30489999055862427},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.2978000044822693},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.29440000653266907},{"id":"https://openalex.org/C177212765","wikidata":"https://www.wikidata.org/wiki/Q627335","display_name":"Workflow","level":2,"score":0.29170000553131104},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.2847000062465668},{"id":"https://openalex.org/C181889124","wikidata":"https://www.wikidata.org/wiki/Q380204","display_name":"Service level","level":2,"score":0.2791999876499176},{"id":"https://openalex.org/C2777904410","wikidata":"https://www.wikidata.org/wiki/Q7397","display_name":"Software","level":2,"score":0.26820001006126404},{"id":"https://openalex.org/C2985973956","wikidata":"https://www.wikidata.org/wiki/Q1617745","display_name":"High energy","level":2,"score":0.2621999979019165},{"id":"https://openalex.org/C99862985","wikidata":"https://www.wikidata.org/wiki/Q10858068","display_name":"Production line","level":2,"score":0.2606000006198883},{"id":"https://openalex.org/C111873713","wikidata":"https://www.wikidata.org/wiki/Q1641413","display_name":"Job scheduler","level":3,"score":0.2572999894618988},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.25609999895095825},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.25099998712539673}],"mesh":[],"locations_count":4,"locations":[{"id":"doi:10.1109/ipdps65963.2026.00078","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ipdps65963.2026.00078","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Parallel and Distributed Processing Symposium (IPDPS)","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2502.18680","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.18680","pdf_url":"https://arxiv.org/pdf/2502.18680","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"pmh:doi:10.48550/arxiv.2502.18680","is_oa":true,"landing_page_url":null,"pdf_url":null,"source":{"id":"https://openalex.org/S4406922384","display_name":"Open MIND","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":"publisher-specific-oa","license_id":"https://openalex.org/licenses/publisher-specific-oa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Article"},{"id":"doi:10.48550/arxiv.2502.18680","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2502.18680","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2502.18680","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2502.18680","pdf_url":"https://arxiv.org/pdf/2502.18680","source":{"id":"https://openalex.org/S4393918464","display_name":"ArXiv.org","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320317220","display_name":"National Energy Research Scientific Computing Center","ror":"https://ror.org/05v3mvq14"},{"id":"https://openalex.org/F4320332359","display_name":"Office of Science","ror":"https://ror.org/00mmn6b08"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"GPGPU-accelerated":[0],"clusters":[1],"and":[2,20,36,93,101,139,151],"supercomputers":[3],"are":[4,149],"central":[5],"to":[6,18,118,128,155],"modern":[7],"high-performance":[8],"computing":[9],"(HPC).":[10],"Over":[11],"the":[12,40,49,94],"past":[13],"decade,":[14],"these":[15,43,116],"systems":[16],"continue":[17],"expand,":[19],"GPUs":[21,89],"now":[22],"expose":[23],"a":[24,80,91,110],"wide":[25],"range":[26],"of":[27,34,42,96,124,147],"hardware":[28,70],"counters":[29,71],"that":[30],"provide":[31,129],"detailed":[32],"views":[33],"performance":[35],"resource":[37],"usage.":[38],"Despite":[39],"potential":[41],"counters,":[44],"few":[45],"studies":[46],"have":[47],"evaluated":[48],"insights":[50],"they":[51],"offer":[52],"about":[53],"real":[54],"workloads":[55],"at":[56,162,173],"scale.":[57],"In":[58],"this":[59,63],"work,":[60],"we":[61,102],"address":[62],"gap":[64],"by":[65],"analyzing":[66],"previously":[67],"underexplored":[68],"GPU":[69,97],"collected":[72],"via":[73],"Lightweight":[74],"Distributed":[75],"Metric":[76],"Service":[77],"on":[78],"Perlmutter,":[79],"leadership-class":[81],"supercomputer.":[82],"We":[83,113],"quantify":[84],"uneven":[85],"work":[86],"distribution":[87],"across":[88],"within":[90],"job":[92,120],"steadiness":[95],"activity":[98],"over":[99],"time,":[100],"classify":[103],"jobs":[104,148,153,161,166],"as":[105],"compute-":[106],"or":[107,177],"memory-bound":[108,152],"using":[109],"roofline-based":[111],"criterion.":[112],"then":[114],"use":[115],"metrics":[117],"interpret":[119],"behavior":[121],"in":[122],"terms":[123],"practical":[125],"workload":[126,137],"characteristics":[127],"interpretable,":[130],"job-level":[131],"insights.":[132],"Our":[133],"findings":[134],"can":[135],"inform":[136],"optimization":[138],"future":[140],"HPC":[141],"system":[142],"design.":[143],"For":[144],"example,":[145],"81%":[146],"memory-bound,":[150],"tend":[154],"consume":[156],"more":[157],"energy":[158],"than":[159],"compute-bound":[160],"comparable":[163],"GPU-hours.":[164],"Among":[165],"requesting":[167],"80":[168],"GB":[169],"GPUs,":[170],"55%":[171],"peak":[172],"50%":[174],"HBM":[175],"capacity":[176],"less.":[178]},"counts_by_year":[],"updated_date":"2026-07-17T09:13:05.818461","created_date":"2025-10-10T00:00:00"}