{"id":"https://openalex.org/W7162457989","doi":"https://doi.org/10.1109/ispass69572.2026.00051","title":"RACER: A Caching Layer based Optimization for Accelerating Reinforcement Learning Workloads","display_name":"RACER: A Caching Layer based Optimization for Accelerating Reinforcement Learning Workloads","publication_year":2026,"publication_date":"2026-04-26","ids":{"openalex":"https://openalex.org/W7162457989","doi":"https://doi.org/10.1109/ispass69572.2026.00051"},"language":null,"primary_location":{"id":"doi:10.1109/ispass69572.2026.00051","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ispass69572.2026.00051","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)","raw_type":"proceedings-article"},"type":"conference-paper","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/A5006897355","display_name":"Kailash Gogineni","orcid":"https://orcid.org/0000-0003-1865-5470"},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kailash Gogineni","raw_affiliation_strings":["George Washington University,USA","Independent Researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]},{"raw_affiliation_string":"Independent Researcher","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011252336","display_name":"Yongsheng Mei","orcid":"https://orcid.org/0000-0001-7606-8931"},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yongsheng Mei","raw_affiliation_strings":["George Washington University,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5097357211","display_name":"Karthikeya Gogineni","orcid":null},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Karthikeya Gogineni","raw_affiliation_strings":["George Washington University,USA","Independent Researcher"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]},{"raw_affiliation_string":"Independent Researcher","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137051603","display_name":"Peng Wei","orcid":null},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Peng Wei","raw_affiliation_strings":["George Washington University,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137069076","display_name":"Tian Lan","orcid":null},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Tian Lan","raw_affiliation_strings":["George Washington University,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5137054069","display_name":"Guru Venkataramani","orcid":null},"institutions":[{"id":"https://openalex.org/I193531525","display_name":"George Washington University","ror":"https://ror.org/00y4zzh67","country_code":"US","type":"education","lineage":["https://openalex.org/I193531525"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Guru Venkataramani","raw_affiliation_strings":["George Washington University,USA"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"George Washington University,USA","institution_ids":["https://openalex.org/I193531525"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":["https://openalex.org/I193531525"],"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":"461","last_page":"472"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.4948999881744385,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.4948999881744385,"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/T11478","display_name":"Caching and Content Delivery","score":0.06920000165700912,"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/T13553","display_name":"Age of Information Optimization","score":0.026399999856948853,"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/layer","display_name":"Layer (electronics)","score":0.5947999954223633},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5685999989509583},{"id":"https://openalex.org/keywords/control","display_name":"Control (management)","score":0.3418000042438507},{"id":"https://openalex.org/keywords/set","display_name":"Set (abstract data type)","score":0.32919999957084656},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.2632000148296356}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6952000260353088},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.5947999954223633},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5685999989509583},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.3418000042438507},{"id":"https://openalex.org/C177264268","wikidata":"https://www.wikidata.org/wiki/Q1514741","display_name":"Set (abstract data type)","level":2,"score":0.32919999957084656},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.30790001153945923},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.29829999804496765},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.29499998688697815},{"id":"https://openalex.org/C31258907","wikidata":"https://www.wikidata.org/wiki/Q1301371","display_name":"Computer network","level":1,"score":0.29339998960494995},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.2632000148296356},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.257999986410141},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.257099986076355},{"id":"https://openalex.org/C190793597","wikidata":"https://www.wikidata.org/wiki/Q189768","display_name":"Application layer","level":3,"score":0.2556999921798706}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/ispass69572.2026.00051","is_oa":false,"landing_page_url":"https://doi.org/10.1109/ispass69572.2026.00051","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2026 IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"},{"id":"https://openalex.org/F4320337345","display_name":"Office of Naval Research","ror":"https://ror.org/00rk2pe57"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":34,"referenced_works":["https://openalex.org/W2002555321","https://openalex.org/W2012465543","https://openalex.org/W2018689653","https://openalex.org/W2791310449","https://openalex.org/W2931767035","https://openalex.org/W2972392561","https://openalex.org/W2980564817","https://openalex.org/W2991431362","https://openalex.org/W3012628688","https://openalex.org/W3043360281","https://openalex.org/W3048346032","https://openalex.org/W3103837983","https://openalex.org/W3135314352","https://openalex.org/W3138303811","https://openalex.org/W3187188899","https://openalex.org/W3200211247","https://openalex.org/W3207399097","https://openalex.org/W3207927582","https://openalex.org/W3212514695","https://openalex.org/W4214700277","https://openalex.org/W4220884018","https://openalex.org/W4252238872","https://openalex.org/W4281657584","https://openalex.org/W4360831955","https://openalex.org/W4383816138","https://openalex.org/W4385568106","https://openalex.org/W4387269743","https://openalex.org/W4393406936","https://openalex.org/W4400680802","https://openalex.org/W4404848788","https://openalex.org/W4411358560","https://openalex.org/W4413412048","https://openalex.org/W7124300950","https://openalex.org/W7133195748"],"related_works":[],"abstract_inverted_index":{"Reinforcement":[0],"Learning":[1],"(RL)":[2],"learns":[3],"optimal":[4],"decision-making":[5],"policies":[6,42],"from":[7,52,65,125,152],"experiential":[8],"(transition)":[9],"datasets":[10],"and":[11,43,81,87,158,179,191,241,266],"maximizes":[12],"the":[13,22,58,74,96,117,139,146,160,184,194,208,222,226,255],"RL":[14,23,46,111,118,235,256],"agent\u2019s":[15],"cumulative":[16],"rewards.":[17,44,276],"In":[18],"order":[19],"to":[20,31,40,72,136,164,170,182,192,200,216,254,271],"improve":[21],"training":[24,98,112,268],"efficiency,":[25],"prior":[26],"works":[27],"have":[28],"studied":[29],"how":[30,169],"selectively":[32],"sample":[33],"certain":[34],"critical":[35,134,156,227],"transitions":[36,135,157,187],"that":[37,116,132,197,220,247],"ultimately":[38],"lead":[39],"better":[41],"However,":[45],"workloads":[47],"still":[48],"face":[49],"significant":[50,250],"challenges":[51],"a":[53,130,153],"systems":[54],"perspective,":[55],"particularly":[56],"when":[57],"agent":[59,119],"iteratively":[60],"accesses":[61,121],"batches":[62,124],"of":[63,155,186,204,224,263],"data":[64,196,258],"transition":[66,123,195,257],"datasets,":[67,127],"whose":[68],"growing":[69,126],"sizes":[70],"continue":[71],"challenge":[73],"memory":[75,83,147,162],"hierarchy.":[76],"This":[77,143],"results":[78],"in":[79],"frequent":[80],"costly":[82,161],"transfers":[84],"between":[85],"caches":[86],"Dynamic":[88],"Random":[89],"Access":[90],"Memory":[91],"(DRAM),":[92],"which":[93],"negatively":[94],"impacts":[95],"overall":[97],"time.In":[99],"this":[100],"paper,":[101],"we":[102,128,167],"propose":[103],"RACER,":[104],"our":[105,217],"novel":[106],"caching":[107,209,218],"layerbased":[108],"optimization":[109,190,251],"for":[110],"workloads.":[113],"Firstly,":[114],"recognizing":[115],"repeatedly":[120],"large":[122],"design":[129,144],"storage-cache":[131],"prioritizes":[133],"fit":[137],"within":[138],"hardware":[140],"cache":[141],"hierarchies.":[142],"reduces":[145],"access":[148],"times":[149],"by":[150],"sampling":[151,259],"subset":[154],"minimizes":[159],"trips":[163],"DRAM.":[165],"Second,":[166],"demonstrate":[168],"smartly":[171],"leverage":[172],"key":[173],"metrics":[174],"(viz.,":[175],"temporal":[176],"difference":[177],"error":[178],"advantage":[180],"weighting)":[181],"quantify":[183],"importance/relevance":[185],"during":[188],"policy":[189],"identify":[193],"would":[198],"need":[199],"be":[201],"spilled":[202],"out":[203],"(or":[205],"filled":[206],"into)":[207],"system.":[210],"We":[211],"also":[212],"introduce":[213],"dynamic":[214],"optimizations":[215],"system":[219],"minimize":[221],"prospect":[223],"discarding":[225],"transitions.":[228],"Our":[229],"performance":[230],"evaluation":[231],"across":[232],"three":[233,243],"state-of-the-art":[234],"algorithms,":[236],"under":[237],"various":[238],"task":[239],"environments,":[240],"on":[242],"different":[244],"systems,":[245],"demonstrates":[246],"RACER":[248],"achieves":[249],"time":[252,269],"improvements":[253],"phase":[260],"(a":[261],"speedup":[262],"$6":[264],"\\times$)":[265,273],"end-to-end":[267],"(up":[270],"$2":[272],"with":[274],"comparable":[275]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2026-05-27T00:00:00"}