{"id":"https://openalex.org/W7170105140","doi":"https://doi.org/10.1145/3785462.3815871","title":"Getting the Most Out of Your GPUs: Automatic Cancellation of Low Efficiency Jobs, GPU Fractionalization, Detailed Metrics, and GPU Sharing","display_name":"Getting the Most Out of Your GPUs: Automatic Cancellation of Low Efficiency Jobs, GPU Fractionalization, Detailed Metrics, and GPU Sharing","publication_year":2026,"publication_date":"2026-07-22","ids":{"openalex":"https://openalex.org/W7170105140","doi":"https://doi.org/10.1145/3785462.3815871"},"language":null,"primary_location":{"id":"doi:10.1145/3785462.3815871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3785462.3815871","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 Practice and Experience in Advanced Research Computing 2026: Resilient Roots + Empowered Communities","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1145/3785462.3815871","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5036623197","display_name":"Jonathan D. Halverson","orcid":"https://orcid.org/0009-0000-4877-3836"},"institutions":[{"id":"https://openalex.org/I20089843","display_name":"Princeton University","ror":"https://ror.org/00hx57361","country_code":"US","type":"education","lineage":["https://openalex.org/I20089843"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jonathan Halverson","raw_affiliation_strings":["Research Computing, Princeton University, Princeton, NJ, USA"],"raw_orcid":"https://orcid.org/0009-0000-4877-3836","affiliations":[{"raw_affiliation_string":"Research Computing, Princeton University, Princeton, NJ, USA","institution_ids":["https://openalex.org/I20089843"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5092841083","display_name":"Josko Plazonic","orcid":"https://orcid.org/0009-0000-6648-3568"},"institutions":[{"id":"https://openalex.org/I20089843","display_name":"Princeton University","ror":"https://ror.org/00hx57361","country_code":"US","type":"education","lineage":["https://openalex.org/I20089843"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Josko Plazonic","raw_affiliation_strings":["Research Computing, Princeton University, Princeton, NJ, USA"],"raw_orcid":"https://orcid.org/0009-0000-6648-3568","affiliations":[{"raw_affiliation_string":"Research Computing, Princeton University, Princeton, NJ, USA","institution_ids":["https://openalex.org/I20089843"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5003806610","display_name":"Johnathan Lee","orcid":"https://orcid.org/0009-0005-1200-1670"},"institutions":[{"id":"https://openalex.org/I55732556","display_name":"Arizona State University","ror":"https://ror.org/03efmqc40","country_code":"US","type":"education","lineage":["https://openalex.org/I55732556"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Johnathan Lee","raw_affiliation_strings":["Research Computing, Arizona State University, Tempe, AZ, USA"],"raw_orcid":"https://orcid.org/0009-0005-1200-1670","affiliations":[{"raw_affiliation_string":"Research Computing, Arizona State University, Tempe, AZ, USA","institution_ids":["https://openalex.org/I55732556"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5056999022","display_name":"Samyak Gupta","orcid":"https://orcid.org/0009-0008-2906-2236"},"institutions":[{"id":"https://openalex.org/I20089843","display_name":"Princeton University","ror":"https://ror.org/00hx57361","country_code":"US","type":"education","lineage":["https://openalex.org/I20089843"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Samyak Gupta","raw_affiliation_strings":["Computer Science, Princeton University, Princeton, NJ, USA"],"raw_orcid":"https://orcid.org/0000-0002-3258-0789","affiliations":[{"raw_affiliation_string":"Computer Science, Princeton University, Princeton, NJ, USA","institution_ids":["https://openalex.org/I20089843"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"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":"1","last_page":"5"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":null,"topics":[],"keywords":[{"id":"https://openalex.org/keywords/graphics","display_name":"Graphics","score":0.6101999878883362},{"id":"https://openalex.org/keywords/general-purpose-computing-on-graphics-processing-units","display_name":"General-purpose computing on graphics processing units","score":0.5871999859809875},{"id":"https://openalex.org/keywords/cuda","display_name":"CUDA","score":0.45559999346733093},{"id":"https://openalex.org/keywords/graphics-processing-unit","display_name":"Graphics processing unit","score":0.4523000121116638},{"id":"https://openalex.org/keywords/gpu-cluster","display_name":"GPU cluster","score":0.35659998655319214},{"id":"https://openalex.org/keywords/cover","display_name":"Cover (algebra)","score":0.33390000462532043},{"id":"https://openalex.org/keywords/matching","display_name":"Matching (statistics)","score":0.3325999975204468}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8097000122070312},{"id":"https://openalex.org/C21442007","wikidata":"https://www.wikidata.org/wiki/Q1027879","display_name":"Graphics","level":2,"score":0.6101999878883362},{"id":"https://openalex.org/C50630238","wikidata":"https://www.wikidata.org/wiki/Q971505","display_name":"General-purpose computing on graphics processing units","level":3,"score":0.5871999859809875},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5475999712944031},{"id":"https://openalex.org/C2778119891","wikidata":"https://www.wikidata.org/wiki/Q477690","display_name":"CUDA","level":2,"score":0.45559999346733093},{"id":"https://openalex.org/C2779851693","wikidata":"https://www.wikidata.org/wiki/Q183484","display_name":"Graphics processing unit","level":2,"score":0.4523000121116638},{"id":"https://openalex.org/C2781335571","wikidata":"https://www.wikidata.org/wiki/Q2633544","display_name":"GPU cluster","level":3,"score":0.35659998655319214},{"id":"https://openalex.org/C2780428219","wikidata":"https://www.wikidata.org/wiki/Q16952335","display_name":"Cover (algebra)","level":2,"score":0.33390000462532043},{"id":"https://openalex.org/C165064840","wikidata":"https://www.wikidata.org/wiki/Q1321061","display_name":"Matching (statistics)","level":2,"score":0.3325999975204468},{"id":"https://openalex.org/C4822641","wikidata":"https://www.wikidata.org/wiki/Q846651","display_name":"Multiprocessing","level":2,"score":0.3248000144958496},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.3237000107765198},{"id":"https://openalex.org/C4924752","wikidata":"https://www.wikidata.org/wiki/Q184148","display_name":"Plug-in","level":2,"score":0.31709998846054077},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.28119999170303345},{"id":"https://openalex.org/C78766204","wikidata":"https://www.wikidata.org/wiki/Q555032","display_name":"Multi-core processor","level":2,"score":0.2802000045776367},{"id":"https://openalex.org/C121684516","wikidata":"https://www.wikidata.org/wiki/Q7600677","display_name":"Computer graphics (images)","level":1,"score":0.26109999418258667},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.25769999623298645},{"id":"https://openalex.org/C2777212361","wikidata":"https://www.wikidata.org/wiki/Q5127848","display_name":"Class (philosophy)","level":2,"score":0.25760000944137573},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.2563999891281128},{"id":"https://openalex.org/C67953723","wikidata":"https://www.wikidata.org/wiki/Q192525","display_name":"Workstation","level":2,"score":0.25450000166893005}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3785462.3815871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3785462.3815871","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 Practice and Experience in Advanced Research Computing 2026: Resilient Roots + Empowered Communities","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3785462.3815871","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3785462.3815871","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 Practice and Experience in Advanced Research Computing 2026: Resilient Roots + Empowered Communities","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":9,"referenced_works":["https://openalex.org/W1596936080","https://openalex.org/W2270461552","https://openalex.org/W2804145951","https://openalex.org/W4289828024","https://openalex.org/W4386585137","https://openalex.org/W4386585189","https://openalex.org/W4396721167","https://openalex.org/W4411620058","https://openalex.org/W4413059339"],"related_works":[],"abstract_inverted_index":{"Graphics":[0],"Processing":[1],"Units":[2],"(GPUs)":[3],"are":[4,46,54,138],"expensive":[5],"and":[6,85,120,127],"in":[7,25,92,96,139],"great":[8],"demand":[9],"at":[10],"nearly":[11],"all":[12,68],"research":[13],"institutions.":[14],"We":[15],"present":[16,74],"four":[17],"ways":[18],"for":[19,35,149],"institutions":[20],"to":[21,129,134],"maximize":[22],"their":[23],"investment":[24],"GPUs":[26,101],"without":[27],"incurring":[28],"any":[29],"new":[30],"financial":[31],"costs.":[32],"First,":[33],"tools":[34,60],"automatically":[36],"cancelling":[37],"Slurm":[38,147],"jobs":[39],"with":[40,61],"zero":[41],"or":[42,83],"low":[43],"GPU":[44,78,82,110,150],"utilization":[45],"discussed.":[47],"Our":[48],"survey":[49],"identified":[50],"nine":[51],"universities":[52],"that":[53,90,132],"currently":[55],"taking":[56],"advantage":[57],"of":[58,63,67,77,100,145],"these":[59],"one":[62,87],"them":[64],"saving":[65],"6%":[66],"GPU-hours":[69],"by":[70],"auto-cancelling.":[71],"Second,":[72],"we":[73,106],"the":[75,97],"merits":[76],"fractionalization":[79],"(e.g.,":[80],"Multi-instance":[81],"MIG)":[84],"cover":[86],"example":[88],"case":[89],"resulted":[91],"a":[93,103,143,146],"15%":[94],"increase":[95],"effective":[98],"number":[99],"on":[102],"cluster.":[104],"Third,":[105],"show":[107],"how":[108],"detailed":[109],"metrics":[111],"such":[112],"as":[113],"streaming":[114],"multiprocessor":[115],"utilization,":[116,119],"Tensor":[117],"Core":[118],"NVLink":[121],"transfer":[122],"rates":[123],"can":[124],"be":[125,135],"measured":[126],"used":[128],"identify":[130],"codes":[131],"appear":[133],"efficient":[136],"but":[137],"fact":[140],"underperforming.":[141],"Lastly,":[142],"preview":[144],"plugin":[148],"sharing":[151],"is":[152],"presented.":[153]},"counts_by_year":[],"updated_date":"2026-07-23T06:02:39.697066","created_date":"2026-07-23T00:00:00"}