{"id":"https://openalex.org/W3180201702","doi":"https://doi.org/10.1145/3458817.3476152","title":"KAISA","display_name":"KAISA","publication_year":2021,"publication_date":"2021-11-13","ids":{"openalex":"https://openalex.org/W3180201702","doi":"https://doi.org/10.1145/3458817.3476152","mag":"3180201702"},"language":"en","primary_location":{"id":"doi:10.1145/3458817.3476152","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3458817.3476152","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2107.01739","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5069913075","display_name":"J. Gregory Pauloski","orcid":"https://orcid.org/0000-0002-6547-6902"},"institutions":[{"id":"https://openalex.org/I40347166","display_name":"University of Chicago","ror":"https://ror.org/024mw5h28","country_code":"US","type":"education","lineage":["https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"J. Gregory Pauloski","raw_affiliation_strings":["University of Chicago"],"affiliations":[{"raw_affiliation_string":"University of Chicago","institution_ids":["https://openalex.org/I40347166"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5087677226","display_name":"Qi Huang","orcid":"https://orcid.org/0000-0002-9425-6618"},"institutions":[{"id":"https://openalex.org/I86519309","display_name":"The University of Texas at Austin","ror":"https://ror.org/00hj54h04","country_code":"US","type":"education","lineage":["https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Qi Huang","raw_affiliation_strings":["University of Texas at Austin"],"affiliations":[{"raw_affiliation_string":"University of Texas at Austin","institution_ids":["https://openalex.org/I86519309"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101670806","display_name":"Lei Huang","orcid":"https://orcid.org/0000-0003-0502-168X"},"institutions":[{"id":"https://openalex.org/I4388891828","display_name":"Texas Advanced Computing Center","ror":"https://ror.org/00xg4bh43","country_code":null,"type":"facility","lineage":["https://openalex.org/I4388891828","https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lei Huang","raw_affiliation_strings":["Texas Advanced Computing Center"],"affiliations":[{"raw_affiliation_string":"Texas Advanced Computing Center","institution_ids":["https://openalex.org/I4388891828"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082710354","display_name":"Shivaram Venkataraman","orcid":"https://orcid.org/0000-0001-9575-7935"},"institutions":[{"id":"https://openalex.org/I135310074","display_name":"University of Wisconsin\u2013Madison","ror":"https://ror.org/01y2jtd41","country_code":"US","type":"education","lineage":["https://openalex.org/I135310074"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Shivaram Venkataraman","raw_affiliation_strings":["University of Wisconsin, Madison"],"affiliations":[{"raw_affiliation_string":"University of Wisconsin, Madison","institution_ids":["https://openalex.org/I135310074"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5065464552","display_name":"Kyle Chard","orcid":"https://orcid.org/0000-0002-7370-4805"},"institutions":[{"id":"https://openalex.org/I40347166","display_name":"University of Chicago","ror":"https://ror.org/024mw5h28","country_code":"US","type":"education","lineage":["https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Kyle Chard","raw_affiliation_strings":["University of Chicago"],"affiliations":[{"raw_affiliation_string":"University of Chicago","institution_ids":["https://openalex.org/I40347166"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5032231503","display_name":"Ian Foster","orcid":"https://orcid.org/0000-0003-2129-5269"},"institutions":[{"id":"https://openalex.org/I40347166","display_name":"University of Chicago","ror":"https://ror.org/024mw5h28","country_code":"US","type":"education","lineage":["https://openalex.org/I40347166"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ian Foster","raw_affiliation_strings":["University of Chicago"],"affiliations":[{"raw_affiliation_string":"University of Chicago","institution_ids":["https://openalex.org/I40347166"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5100423029","display_name":"Zhao Zhang","orcid":"https://orcid.org/0000-0002-5703-7969"},"institutions":[{"id":"https://openalex.org/I4388891828","display_name":"Texas Advanced Computing Center","ror":"https://ror.org/00xg4bh43","country_code":null,"type":"facility","lineage":["https://openalex.org/I4388891828","https://openalex.org/I86519309"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zhao Zhang","raw_affiliation_strings":["Texas Advanced Computing Center"],"affiliations":[{"raw_affiliation_string":"Texas Advanced Computing Center","institution_ids":["https://openalex.org/I4388891828"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5069913075"],"corresponding_institution_ids":["https://openalex.org/I40347166"],"apc_list":null,"apc_paid":null,"fwci":0.7751,"has_fulltext":false,"cited_by_count":10,"citation_normalized_percentile":{"value":0.73593867,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":94,"max":97},"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"14"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10036","display_name":"Advanced Neural Network Applications","score":1.0,"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":1.0,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.9991999864578247,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.998199999332428,"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/memory-footprint","display_name":"Memory footprint","score":0.9027879238128662},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.8227869272232056},{"id":"https://openalex.org/keywords/scalability","display_name":"Scalability","score":0.8099329471588135},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.5942206382751465},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.5801618099212646},{"id":"https://openalex.org/keywords/stochastic-gradient-descent","display_name":"Stochastic gradient descent","score":0.4912685751914978},{"id":"https://openalex.org/keywords/throughput","display_name":"Throughput","score":0.482399582862854},{"id":"https://openalex.org/keywords/residual-neural-network","display_name":"Residual neural network","score":0.4741196930408478},{"id":"https://openalex.org/keywords/scaling","display_name":"Scaling","score":0.4129437208175659},{"id":"https://openalex.org/keywords/computer-engineering","display_name":"Computer engineering","score":0.4000844657421112},{"id":"https://openalex.org/keywords/artificial-neural-network","display_name":"Artificial neural network","score":0.36879390478134155},{"id":"https://openalex.org/keywords/algorithm","display_name":"Algorithm","score":0.31961745023727417},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.2036929726600647},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.10665121674537659}],"concepts":[{"id":"https://openalex.org/C74912251","wikidata":"https://www.wikidata.org/wiki/Q6815727","display_name":"Memory footprint","level":2,"score":0.9027879238128662},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8227869272232056},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.8099329471588135},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.5942206382751465},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.5801618099212646},{"id":"https://openalex.org/C206688291","wikidata":"https://www.wikidata.org/wiki/Q7617819","display_name":"Stochastic gradient descent","level":3,"score":0.4912685751914978},{"id":"https://openalex.org/C157764524","wikidata":"https://www.wikidata.org/wiki/Q1383412","display_name":"Throughput","level":3,"score":0.482399582862854},{"id":"https://openalex.org/C2944601119","wikidata":"https://www.wikidata.org/wiki/Q43744058","display_name":"Residual neural network","level":3,"score":0.4741196930408478},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.4129437208175659},{"id":"https://openalex.org/C113775141","wikidata":"https://www.wikidata.org/wiki/Q428691","display_name":"Computer engineering","level":1,"score":0.4000844657421112},{"id":"https://openalex.org/C50644808","wikidata":"https://www.wikidata.org/wiki/Q192776","display_name":"Artificial neural network","level":2,"score":0.36879390478134155},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.31961745023727417},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.2036929726600647},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.10665121674537659},{"id":"https://openalex.org/C555944384","wikidata":"https://www.wikidata.org/wiki/Q249","display_name":"Wireless","level":2,"score":0.0},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C76155785","wikidata":"https://www.wikidata.org/wiki/Q418","display_name":"Telecommunications","level":1,"score":0.0},{"id":"https://openalex.org/C111919701","wikidata":"https://www.wikidata.org/wiki/Q9135","display_name":"Operating system","level":1,"score":0.0},{"id":"https://openalex.org/C77088390","wikidata":"https://www.wikidata.org/wiki/Q8513","display_name":"Database","level":1,"score":0.0}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3458817.3476152","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3458817.3476152","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2107.01739","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2107.01739","pdf_url":"https://arxiv.org/pdf/2107.01739","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2107.01739","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2107.01739","pdf_url":"https://arxiv.org/pdf/2107.01739","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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":76,"referenced_works":["https://openalex.org/W95608104","https://openalex.org/W201315547","https://openalex.org/W1566289585","https://openalex.org/W1861492603","https://openalex.org/W1901129140","https://openalex.org/W2051434435","https://openalex.org/W2060393849","https://openalex.org/W2083842231","https://openalex.org/W2103111465","https://openalex.org/W2110873104","https://openalex.org/W2138243089","https://openalex.org/W2155894447","https://openalex.org/W2163605009","https://openalex.org/W2186615578","https://openalex.org/W2194775991","https://openalex.org/W2271349422","https://openalex.org/W2340006107","https://openalex.org/W2402144811","https://openalex.org/W2525778437","https://openalex.org/W2604272474","https://openalex.org/W2750933313","https://openalex.org/W2763421725","https://openalex.org/W2774000609","https://openalex.org/W2779888504","https://openalex.org/W2786263712","https://openalex.org/W2787998955","https://openalex.org/W2806659157","https://openalex.org/W2896457183","https://openalex.org/W2899071864","https://openalex.org/W2902363477","https://openalex.org/W2945554113","https://openalex.org/W2945785363","https://openalex.org/W2951781666","https://openalex.org/W2952122856","https://openalex.org/W2953384591","https://openalex.org/W2962739609","https://openalex.org/W2962883024","https://openalex.org/W2963150697","https://openalex.org/W2963341956","https://openalex.org/W2963344792","https://openalex.org/W2963390429","https://openalex.org/W2963408258","https://openalex.org/W2963433607","https://openalex.org/W2963446712","https://openalex.org/W2963803379","https://openalex.org/W2964108773","https://openalex.org/W2967791890","https://openalex.org/W2969388332","https://openalex.org/W2970971581","https://openalex.org/W2971055146","https://openalex.org/W2971890617","https://openalex.org/W2974008169","https://openalex.org/W2991040477","https://openalex.org/W2995435108","https://openalex.org/W3005522790","https://openalex.org/W3012479151","https://openalex.org/W3030163527","https://openalex.org/W3035182906","https://openalex.org/W3037639655","https://openalex.org/W3080999573","https://openalex.org/W3081233814","https://openalex.org/W3096403968","https://openalex.org/W3096583839","https://openalex.org/W3096956001","https://openalex.org/W3098238825","https://openalex.org/W3099472489","https://openalex.org/W3103894541","https://openalex.org/W3131087743","https://openalex.org/W4212774754","https://openalex.org/W4292779060","https://openalex.org/W4295181861","https://openalex.org/W4295312788","https://openalex.org/W4295830359","https://openalex.org/W4297808460","https://openalex.org/W4301239768","https://openalex.org/W4301361180"],"related_works":["https://openalex.org/W2599472179","https://openalex.org/W4323057981","https://openalex.org/W3178607569","https://openalex.org/W4310870954","https://openalex.org/W4308408209","https://openalex.org/W4386900849","https://openalex.org/W2893646015","https://openalex.org/W4396520111","https://openalex.org/W2929170389","https://openalex.org/W4300097863"],"abstract_inverted_index":{"Kronecker-factored":[0],"Approximate":[1],"Curvature":[2],"(K-FAC)":[3],"has":[4],"recently":[5],"been":[6],"shown":[7],"to":[8,30,58,88,94,132,137],"converge":[9],"faster":[10,101,120],"in":[11,121],"deep":[12],"neural":[13],"network":[14],"(DNN)":[15],"training":[16],"than":[17,140],"stochastic":[18],"gradient":[19],"descent":[20],"(SGD);":[21],"however,":[22],"K-FAC's":[23],"larger":[24],"memory":[25,48,69,113,129],"footprint":[26],"hinders":[27],"its":[28],"applicability":[29],"large":[31,77],"models.":[32],"We":[33,64],"present":[34],"KAISA,":[35],"a":[36,111],"K-FAC-enabled,":[37],"Adaptable,":[38],"Improved,":[39],"and":[40,51,56,61,70,73,84,118,123,130],"ScAlable":[41],"second-order":[42],"optimizer":[43],"framework":[44],"that":[45],"adapts":[46],"the":[47,66,95,105,141],"footprint,":[49],"communication,":[50],"computation":[52],"given":[53],"specific":[54],"models":[55],"hardware":[57],"improve":[59],"performance":[60],"increase":[62],"scalability.":[63],"quantify":[65],"tradeoffs":[67],"between":[68],"communication":[71,131],"cost":[72],"evaluate":[74],"KAISA":[75,98,115,126],"on":[76,86],"models,":[78],"including":[79],"ResNet-50,":[80],"Mask":[81],"R-CNN,":[82],"U-Net,":[83],"BERT,":[85],"up":[87],"128":[89],"NVIDIA":[90],"A100":[91],"GPUs.":[92],"Compared":[93],"original":[96],"optimizers,":[97],"converges":[99,116],"18.1--36.3%":[100],"across":[102],"applications":[103],"with":[104],"same":[106],"global":[107],"batch":[108],"size.":[109],"Under":[110],"fixed":[112],"budget,":[114],"32.5%":[117],"41.6%":[119],"ResNet-50":[122],"BERT-Large,":[124],"respectively.":[125],"can":[127],"balance":[128],"achieve":[133],"scaling":[134],"efficiency":[135],"equal":[136],"or":[138],"better":[139],"baseline":[142],"optimizers.":[143]},"counts_by_year":[{"year":2025,"cited_by_count":2},{"year":2024,"cited_by_count":2},{"year":2023,"cited_by_count":3},{"year":2022,"cited_by_count":3}],"updated_date":"2026-03-20T23:20:44.827607","created_date":"2021-07-19T00:00:00"}