{"id":"https://openalex.org/W7124162084","doi":"https://doi.org/10.1109/icpads67057.2025.11323016","title":"QPO: Accelerating Memory-Efficient DNN Training with Quantization and Pipelining","display_name":"QPO: Accelerating Memory-Efficient DNN Training with Quantization and Pipelining","publication_year":2025,"publication_date":"2025-12-14","ids":{"openalex":"https://openalex.org/W7124162084","doi":"https://doi.org/10.1109/icpads67057.2025.11323016"},"language":null,"primary_location":{"id":"doi:10.1109/icpads67057.2025.11323016","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icpads67057.2025.11323016","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS)","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/A5054389074","display_name":"Xiang Fan","orcid":"https://orcid.org/0000-0002-3596-5529"},"institutions":[{"id":"https://openalex.org/I158809036","display_name":"Shenzhen Institute of Information Technology","ror":"https://ror.org/03wrf9427","country_code":"CN","type":"education","lineage":["https://openalex.org/I158809036"]},{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]},{"id":"https://openalex.org/I4210152380","display_name":"Shenzhen Technology University","ror":"https://ror.org/04qzpec27","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210152380"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xiang Fan","raw_affiliation_strings":["School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China","institution_ids":["https://openalex.org/I158809036","https://openalex.org/I204983213","https://openalex.org/I4210152380"]}]},{"author_position":"last","author":{"id":null,"display_name":"Shaohuai Shi","orcid":null},"institutions":[{"id":"https://openalex.org/I158809036","display_name":"Shenzhen Institute of Information Technology","ror":"https://ror.org/03wrf9427","country_code":"CN","type":"education","lineage":["https://openalex.org/I158809036"]},{"id":"https://openalex.org/I204983213","display_name":"Harbin Institute of Technology","ror":"https://ror.org/01yqg2h08","country_code":"CN","type":"education","lineage":["https://openalex.org/I204983213"]},{"id":"https://openalex.org/I4210152380","display_name":"Shenzhen Technology University","ror":"https://ror.org/04qzpec27","country_code":"CN","type":"education","lineage":["https://openalex.org/I4210152380"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Shaohuai Shi","raw_affiliation_strings":["School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"School of Computer Science and Technology, Harbin Institute of Technology,Shenzhen,Shenzhen,China","institution_ids":["https://openalex.org/I158809036","https://openalex.org/I204983213","https://openalex.org/I4210152380"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.57080448,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":0.5259000062942505,"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":0.5259000062942505,"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/T10273","display_name":"IoT and Edge/Fog Computing","score":0.12300000339746475,"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/T14347","display_name":"Big Data and Digital Economy","score":0.05660000070929527,"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/quantization","display_name":"Quantization (signal processing)","score":0.7283999919891357},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.7282000184059143},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6317999958992004},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.4706000089645386},{"id":"https://openalex.org/keywords/data-transmission","display_name":"Data transmission","score":0.46309998631477356},{"id":"https://openalex.org/keywords/transfer","display_name":"Transfer (computing)","score":0.4120999872684479},{"id":"https://openalex.org/keywords/sequence","display_name":"Sequence (biology)","score":0.37229999899864197},{"id":"https://openalex.org/keywords/direct-memory-access","display_name":"Direct memory access","score":0.36480000615119934}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8813999891281128},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.7283999919891357},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.7282000184059143},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6317999958992004},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.48080000281333923},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.4706000089645386},{"id":"https://openalex.org/C557945733","wikidata":"https://www.wikidata.org/wiki/Q389772","display_name":"Data transmission","level":2,"score":0.46309998631477356},{"id":"https://openalex.org/C2776175482","wikidata":"https://www.wikidata.org/wiki/Q1195816","display_name":"Transfer (computing)","level":2,"score":0.4120999872684479},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.37229999899864197},{"id":"https://openalex.org/C37724790","wikidata":"https://www.wikidata.org/wiki/Q210813","display_name":"Direct memory access","level":3,"score":0.36480000615119934},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.3529999852180481},{"id":"https://openalex.org/C2984842247","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep neural networks","level":3,"score":0.3515999913215637},{"id":"https://openalex.org/C9390403","wikidata":"https://www.wikidata.org/wiki/Q3966","display_name":"Computer hardware","level":1,"score":0.3449999988079071},{"id":"https://openalex.org/C150899416","wikidata":"https://www.wikidata.org/wiki/Q1820378","display_name":"Transfer of learning","level":2,"score":0.3434999883174896},{"id":"https://openalex.org/C86111242","wikidata":"https://www.wikidata.org/wiki/Q859595","display_name":"Coprocessor","level":2,"score":0.3353999853134155},{"id":"https://openalex.org/C176649486","wikidata":"https://www.wikidata.org/wiki/Q2308807","display_name":"Memory management","level":3,"score":0.33219999074935913},{"id":"https://openalex.org/C82687282","wikidata":"https://www.wikidata.org/wiki/Q66221","display_name":"Auxiliary memory","level":2,"score":0.3125999867916107},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.3116999864578247},{"id":"https://openalex.org/C149635348","wikidata":"https://www.wikidata.org/wiki/Q193040","display_name":"Embedded system","level":1,"score":0.3102000057697296},{"id":"https://openalex.org/C79403827","wikidata":"https://www.wikidata.org/wiki/Q3988","display_name":"Real-time computing","level":1,"score":0.3100999891757965},{"id":"https://openalex.org/C71901391","wikidata":"https://www.wikidata.org/wiki/Q7126699","display_name":"Upload","level":2,"score":0.30720001459121704},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.29989999532699585},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.29319998621940613},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.27900001406669617},{"id":"https://openalex.org/C194739806","wikidata":"https://www.wikidata.org/wiki/Q66221","display_name":"Computer data storage","level":2,"score":0.27059999108314514},{"id":"https://openalex.org/C12186640","wikidata":"https://www.wikidata.org/wiki/Q6815743","display_name":"Memory model","level":3,"score":0.25929999351501465},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.258899986743927},{"id":"https://openalex.org/C163258240","wikidata":"https://www.wikidata.org/wiki/Q25342","display_name":"Power (physics)","level":2,"score":0.25859999656677246}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/icpads67057.2025.11323016","is_oa":false,"landing_page_url":"https://doi.org/10.1109/icpads67057.2025.11323016","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"2025 IEEE 31th International Conference on Parallel and Distributed Systems (ICPADS)","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G1417173225","display_name":null,"funder_award_id":"KQTD20240729102154066,KJZD20240903104103005,KJZD20230923115113026","funder_id":"https://openalex.org/F4320336606","funder_display_name":"National Transgenic Science and Technology Program"}],"funders":[{"id":"https://openalex.org/F4320336606","display_name":"National Transgenic Science and Technology Program","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":10,"referenced_works":["https://openalex.org/W2489529491","https://openalex.org/W2798707674","https://openalex.org/W2883283076","https://openalex.org/W2891993230","https://openalex.org/W2962821792","https://openalex.org/W3012514909","https://openalex.org/W3043443960","https://openalex.org/W3081168214","https://openalex.org/W4301361180","https://openalex.org/W4327545654"],"related_works":[],"abstract_inverted_index":{"Deep":[0],"neural":[1],"network":[2],"(DNN)":[3],"training":[4,53],"demands":[5],"significant":[6,88],"computational":[7],"power":[8],"and":[9,23,35,44,83,99,117,137,142],"also":[10],"necessitates":[11],"costly":[12],"device":[13,33,71,78],"memory":[14,72,75,156],"for":[15],"the":[16,29,41,60,70,81,106,114,131],"storage":[17],"of":[18,32,59,122,127,149],"parameters,":[19],"gradients,":[20],"optimizer":[21],"states,":[22],"activations.":[24],"Activations":[25],"generally":[26],"take":[27],"up":[28,150],"largest":[30],"portion":[31],"memory,":[34],"their":[36],"usage":[37],"increases":[38],"linearly":[39],"with":[40,109,124],"mini-batch":[42],"size":[43],"sequence":[45],"length,":[46],"which":[47,102],"are":[48],"two":[49],"main":[50],"hyper-parameters":[51],"in":[52],"language":[54],"models.":[55],"Offloading":[56],"is":[57],"one":[58],"widely":[61],"used":[62],"memory-efficient":[63],"techniques":[64],"by":[65],"transferring":[66],"temporarily":[67],"data":[68,89,108],"from":[69],"to":[73,76,112,129,151],"CPU":[74],"save":[77],"memory.":[79],"However,":[80],"offloading":[82,160],"uploading":[84],"operations":[85],"easily":[86],"cause":[87],"transfer":[90,115],"overheads.":[91],"In":[92],"this":[93],"paper,":[94],"we":[95],"propose":[96],"QPO":[97],"(Quantized":[98],"Pipelined":[100],"Offloading),":[101],"combines":[103],"1)":[104],"compressing":[105],"activation":[107],"low-bit":[110],"quantization":[111],"alleviate":[113],"overhead,":[116],"2)":[118],"pipelining":[119],"communication":[120],"tasks":[121,126],"offloading/uploading":[123],"computation":[125],"feed-forward/backpropagation":[128],"reduce":[130],"iteration":[132],"time.":[133],"Experiments":[134],"on":[135],"GPT-2":[136],"LLaMA-2":[138],"models":[139],"using":[140],"A100":[141],"RTX":[143],"3090":[144],"GPUs":[145],"show":[146],"a":[147],"speedup":[148],"15%":[152],"while":[153],"approaching":[154],"minimal":[155],"requirements":[157],"over":[158],"existing":[159],"approaches.":[161]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-01-15T00:00:00"}