{"id":"https://openalex.org/W4281787632","doi":"https://doi.org/10.1145/3534056.3534935","title":"Overflowing emerging neural network inference tasks from the GPU to the CPU on heterogeneous servers","display_name":"Overflowing emerging neural network inference tasks from the GPU to the CPU on heterogeneous servers","publication_year":2022,"publication_date":"2022-06-06","ids":{"openalex":"https://openalex.org/W4281787632","doi":"https://doi.org/10.1145/3534056.3534935"},"language":"en","primary_location":{"id":"doi:10.1145/3534056.3534935","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3534056.3534935","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 15th ACM International Conference on Systems and Storage","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/A5018757871","display_name":"Adithya Kumar","orcid":"https://orcid.org/0009-0005-3876-0624"},"institutions":[{"id":"https://openalex.org/I130769515","display_name":"Pennsylvania State University","ror":"https://ror.org/04p491231","country_code":"US","type":"education","lineage":["https://openalex.org/I130769515"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Adithya Kumar","raw_affiliation_strings":["The Pennsylvania State University"],"affiliations":[{"raw_affiliation_string":"The Pennsylvania State University","institution_ids":["https://openalex.org/I130769515"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5033253809","display_name":"Anand Sivasubramaniam","orcid":"https://orcid.org/0000-0001-6173-687X"},"institutions":[{"id":"https://openalex.org/I130769515","display_name":"Pennsylvania State University","ror":"https://ror.org/04p491231","country_code":"US","type":"education","lineage":["https://openalex.org/I130769515"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Anand Sivasubramaniam","raw_affiliation_strings":["The Pennsylvania State University"],"affiliations":[{"raw_affiliation_string":"The Pennsylvania State University","institution_ids":["https://openalex.org/I130769515"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5024530225","display_name":"Timothy Zhu","orcid":"https://orcid.org/0000-0001-8394-8953"},"institutions":[{"id":"https://openalex.org/I130769515","display_name":"Pennsylvania State University","ror":"https://ror.org/04p491231","country_code":"US","type":"education","lineage":["https://openalex.org/I130769515"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Timothy Zhu","raw_affiliation_strings":["The Pennsylvania State University"],"affiliations":[{"raw_affiliation_string":"The Pennsylvania State University","institution_ids":["https://openalex.org/I130769515"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":["https://openalex.org/A5018757871"],"corresponding_institution_ids":["https://openalex.org/I130769515"],"apc_list":null,"apc_paid":null,"fwci":0.4026,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.58163589,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":89,"max":96},"biblio":{"volume":null,"issue":null,"first_page":"26","last_page":"39"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10036","display_name":"Advanced Neural Network Applications","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9988999962806702,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.994700014591217,"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/computer-science","display_name":"Computer science","score":0.8654070496559143},{"id":"https://openalex.org/keywords/server","display_name":"Server","score":0.6329872012138367},{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.5355244874954224},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.4401712417602539},{"id":"https://openalex.org/keywords/central-processing-unit","display_name":"Central processing unit","score":0.43253475427627563},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.4144173860549927},{"id":"https://openalex.org/keywords/operating-system","display_name":"Operating system","score":0.26379290223121643},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.17426297068595886}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8654070496559143},{"id":"https://openalex.org/C93996380","wikidata":"https://www.wikidata.org/wiki/Q44127","display_name":"Server","level":2,"score":0.6329872012138367},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.5355244874954224},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.4401712417602539},{"id":"https://openalex.org/C49154492","wikidata":"https://www.wikidata.org/wiki/Q5300","display_name":"Central processing unit","level":2,"score":0.43253475427627563},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.4144173860549927},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.26379290223121643},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.17426297068595886},{"id":"https://openalex.org/C162324750","wikidata":"https://www.wikidata.org/wiki/Q8134","display_name":"Economics","level":0,"score":0.0},{"id":"https://openalex.org/C187736073","wikidata":"https://www.wikidata.org/wiki/Q2920921","display_name":"Management","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3534056.3534935","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3534056.3534935","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 15th ACM International Conference on Systems and Storage","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":32,"referenced_works":["https://openalex.org/W1861492603","https://openalex.org/W2065689629","https://openalex.org/W2121893797","https://openalex.org/W2146757372","https://openalex.org/W2169049902","https://openalex.org/W2171226522","https://openalex.org/W2194775991","https://openalex.org/W2442974303","https://openalex.org/W2559655401","https://openalex.org/W2565639579","https://openalex.org/W2606722458","https://openalex.org/W2734941459","https://openalex.org/W2767236912","https://openalex.org/W2794670651","https://openalex.org/W2798956872","https://openalex.org/W2883830791","https://openalex.org/W2903278032","https://openalex.org/W2941938531","https://openalex.org/W2970971581","https://openalex.org/W2981758446","https://openalex.org/W2993364236","https://openalex.org/W3012249773","https://openalex.org/W3012479151","https://openalex.org/W3012514909","https://openalex.org/W3096858590","https://openalex.org/W3097411828","https://openalex.org/W3099464315","https://openalex.org/W3101104221","https://openalex.org/W3139274998","https://openalex.org/W3159576817","https://openalex.org/W3167720200","https://openalex.org/W6713134421"],"related_works":["https://openalex.org/W2092530219","https://openalex.org/W2388464034","https://openalex.org/W2533125852","https://openalex.org/W2140460949","https://openalex.org/W2105580438","https://openalex.org/W2057435755","https://openalex.org/W3121798572","https://openalex.org/W2018782216","https://openalex.org/W2101552690","https://openalex.org/W2125623414"],"abstract_inverted_index":{"While":[0,154],"current":[1],"deep":[2],"learning":[3],"(DL)":[4],"inference":[5],"runtime":[6,191],"systems":[7],"sequentially":[8],"offload":[9],"the":[10,34,41,83,93,97,101,113,117,127,147,197],"model's":[11],"tasks":[12,32,81,150],"on":[13,19,186,194],"to":[14,33,65,82,132,151,161,180],"an":[15],"available":[16],"GPU/accelerator":[17],"based":[18],"its":[20,124],"capability,":[21],"we":[22,85,170],"make":[23],"a":[24,133,166,172,187],"case":[25],"for":[26,51,103,175,221],"selectively":[27],"redirecting":[28],"some":[29,78],"of":[30,73,79,96,112,149,168,196],"these":[31,80,144],"CPU":[35],"and":[36,107,137,159,201,213],"running":[37],"them":[38],"concurrently":[39],"with":[40],"GPU":[42,102,179,218],"doing":[43],"other":[44],"work.":[45],"This":[46],"new":[47],"opportunity":[48],"specifically":[49],"arises":[50],"emerging":[52,222],"DL":[53],"models":[54],"whose":[55],"data":[56,142],"flow":[57],"graphs":[58],"(DFGs)":[59],"have":[60],"much":[61],"wider":[62],"fan-outs":[63],"compared":[64],"traditional":[66],"ones":[67],"which":[68],"are":[69],"invariably":[70],"linear":[71],"chains":[72],"tasks.":[74],"By":[75],"opportunistically":[76],"moving":[77],"CPU,":[84],"can":[86],"(i)":[87],"shave":[88],"off":[89],"service":[90],"times":[91],"from":[92,178],"critical":[94],"path":[95],"DFG,":[98,128],"(ii)":[99],"devote":[100],"more":[104],"deserving":[105],"tasks,":[106],"(iii)":[108],"improve":[109],"overall":[110],"utilization":[111],"provisioned":[114],"hardware":[115,135],"in":[116,126,139,206,211,217],"server.":[118],"However,":[119],"several":[120],"factors":[121],"such":[122],"as":[123],"criticality":[125],"slowdown":[129],"when":[130],"moved":[131],"different":[134],"engine,":[136],"overheads":[138],"transferring":[140],"input/output":[141],"across":[143],"engines,":[145],"determine":[146],"what/when/how":[148],"be":[152,162],"directed.":[153],"this":[155,184],"is":[156],"computationally":[157],"demanding":[158],"slow":[160],"solved":[163],"optimally,":[164],"through":[165],"series":[167],"rationales":[169],"derive":[171],"fast":[173],"technique":[174,185],"task":[176],"overflow":[177],"CPU.":[181],"We":[182],"implement":[183],"nimble":[188],"heterogeneous":[189],"concurrent":[190],"engine":[192,200],"built":[193],"top":[195],"state-of-the-art":[198],"ONNXRuntime":[199],"demonstrate":[202],">":[203,208,214],"10%":[204],"reduction":[205],"latency,":[207],"19%":[209],"gain":[210],"throughput,":[212],"9.8%":[215],"savings":[216],"memory":[219],"usage":[220],"neural":[223],"network":[224],"models.":[225]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":1},{"year":2023,"cited_by_count":1}],"updated_date":"2025-11-06T03:46:38.306776","created_date":"2025-10-10T00:00:00"}