{"id":"https://openalex.org/W4367859812","doi":"https://doi.org/10.1145/3577193.3593715","title":"Anatomy of High-Performance GEMM with Online Fault Tolerance on GPUs","display_name":"Anatomy of High-Performance GEMM with Online Fault Tolerance on GPUs","publication_year":2023,"publication_date":"2023-06-20","ids":{"openalex":"https://openalex.org/W4367859812","doi":"https://doi.org/10.1145/3577193.3593715"},"language":"en","primary_location":{"id":"doi:10.1145/3577193.3593715","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3577193.3593715","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3577193.3593715","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 37th International Conference on Supercomputing","raw_type":"proceedings-article"},"type":"preprint","indexed_in":["arxiv","crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3577193.3593715","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5075806978","display_name":"Shixun Wu","orcid":"https://orcid.org/0000-0001-8637-3307"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":true,"raw_author_name":"Shixun Wu","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5035339023","display_name":"Yujia Zhai","orcid":"https://orcid.org/0000-0002-2688-8058"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Yujia Zhai","raw_affiliation_strings":["University of California, Riverside, Riverside, California, United States of America"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, United States of America","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100705519","display_name":"Jinyang Liu","orcid":"https://orcid.org/0000-0003-0177-502X"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jinyang Liu","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100779661","display_name":"Jiajun Huang","orcid":"https://orcid.org/0000-0001-5092-3987"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Jiajun Huang","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071045650","display_name":"Zizhe Jian","orcid":"https://orcid.org/0009-0005-2079-8130"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zizhe Jian","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5022569738","display_name":"Bryan M. Wong","orcid":"https://orcid.org/0000-0002-3477-8043"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Bryan Wong","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5061737717","display_name":"Zizhong Chen","orcid":"https://orcid.org/0000-0003-2578-4940"},"institutions":[{"id":"https://openalex.org/I103635307","display_name":"University of California, Riverside","ror":"https://ror.org/03nawhv43","country_code":"US","type":"education","lineage":["https://openalex.org/I103635307"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Zizhong Chen","raw_affiliation_strings":["University of California, Riverside, Riverside, California, USA"],"affiliations":[{"raw_affiliation_string":"University of California, Riverside, Riverside, California, USA","institution_ids":["https://openalex.org/I103635307"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":7,"corresponding_author_ids":["https://openalex.org/A5075806978"],"corresponding_institution_ids":["https://openalex.org/I103635307"],"apc_list":null,"apc_paid":null,"fwci":3.3275,"has_fulltext":true,"cited_by_count":11,"citation_normalized_percentile":{"value":0.92619303,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":94,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"360","last_page":"372"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9997000098228455,"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"}},"topics":[{"id":"https://openalex.org/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9997000098228455,"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/T11005","display_name":"Radiation Effects in Electronics","score":0.9993000030517578,"subfield":{"id":"https://openalex.org/subfields/2208","display_name":"Electrical and Electronic Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10829","display_name":"Interconnection Networks and Systems","score":0.9983999729156494,"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/computer-science","display_name":"Computer science","score":0.8388895988464355},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.7369570732116699},{"id":"https://openalex.org/keywords/fault-tolerance","display_name":"Fault tolerance","score":0.5904616713523865},{"id":"https://openalex.org/keywords/speedup","display_name":"Speedup","score":0.5235761404037476},{"id":"https://openalex.org/keywords/supercomputer","display_name":"Supercomputer","score":0.46430662274360657},{"id":"https://openalex.org/keywords/distributed-computing","display_name":"Distributed computing","score":0.19122454524040222}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8388895988464355},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.7369570732116699},{"id":"https://openalex.org/C63540848","wikidata":"https://www.wikidata.org/wiki/Q3140932","display_name":"Fault tolerance","level":2,"score":0.5904616713523865},{"id":"https://openalex.org/C68339613","wikidata":"https://www.wikidata.org/wiki/Q1549489","display_name":"Speedup","level":2,"score":0.5235761404037476},{"id":"https://openalex.org/C83283714","wikidata":"https://www.wikidata.org/wiki/Q121117","display_name":"Supercomputer","level":2,"score":0.46430662274360657},{"id":"https://openalex.org/C120314980","wikidata":"https://www.wikidata.org/wiki/Q180634","display_name":"Distributed computing","level":1,"score":0.19122454524040222}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1145/3577193.3593715","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3577193.3593715","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3577193.3593715","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 37th International Conference on Supercomputing","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2305.01024","is_oa":true,"landing_page_url":"http://arxiv.org/abs/2305.01024","pdf_url":"https://arxiv.org/pdf/2305.01024","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":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"}],"best_oa_location":{"id":"doi:10.1145/3577193.3593715","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3577193.3593715","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3577193.3593715","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 37th International Conference on Supercomputing","raw_type":"proceedings-article"},"sustainable_development_goals":[{"score":0.8299999833106995,"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16"}],"awards":[{"id":"https://openalex.org/G6078133158","display_name":null,"funder_award_id":"DE-SC0022209","funder_id":"https://openalex.org/F4320306084","funder_display_name":"U.S. Department of Energy"}],"funders":[{"id":"https://openalex.org/F4320306084","display_name":"U.S. Department of Energy","ror":"https://ror.org/01bj3aw27"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4367859812.pdf","grobid_xml":"https://content.openalex.org/works/W4367859812.grobid-xml"},"referenced_works_count":58,"referenced_works":["https://openalex.org/W1541483005","https://openalex.org/W1898823215","https://openalex.org/W1986905947","https://openalex.org/W1988070283","https://openalex.org/W1995746640","https://openalex.org/W1997843200","https://openalex.org/W2023856022","https://openalex.org/W2030806070","https://openalex.org/W2034593585","https://openalex.org/W2052455844","https://openalex.org/W2083613288","https://openalex.org/W2095928739","https://openalex.org/W2105524676","https://openalex.org/W2110701642","https://openalex.org/W2118832200","https://openalex.org/W2125768532","https://openalex.org/W2128511938","https://openalex.org/W2130076536","https://openalex.org/W2130189691","https://openalex.org/W2132572492","https://openalex.org/W2134320686","https://openalex.org/W2150326063","https://openalex.org/W2150981663","https://openalex.org/W2151984682","https://openalex.org/W2152211247","https://openalex.org/W2155893237","https://openalex.org/W2156514327","https://openalex.org/W2169596872","https://openalex.org/W2170196949","https://openalex.org/W2229245554","https://openalex.org/W2292469857","https://openalex.org/W2296204683","https://openalex.org/W2343351966","https://openalex.org/W2411755313","https://openalex.org/W2418331349","https://openalex.org/W2647773517","https://openalex.org/W2767260595","https://openalex.org/W2767321582","https://openalex.org/W2767694495","https://openalex.org/W2798289565","https://openalex.org/W2804032941","https://openalex.org/W2950094539","https://openalex.org/W2962821792","https://openalex.org/W2986161099","https://openalex.org/W3014034225","https://openalex.org/W3014127403","https://openalex.org/W3105862567","https://openalex.org/W3140366478","https://openalex.org/W3203693205","https://openalex.org/W3206892724","https://openalex.org/W4210781224","https://openalex.org/W4233556486","https://openalex.org/W4249693612","https://openalex.org/W4256701974","https://openalex.org/W4281758439","https://openalex.org/W4295312788","https://openalex.org/W4383878781","https://openalex.org/W4384705353"],"related_works":["https://openalex.org/W2317245370","https://openalex.org/W4249323025","https://openalex.org/W198851386","https://openalex.org/W2030310580","https://openalex.org/W947442053","https://openalex.org/W1980160788","https://openalex.org/W2148915962","https://openalex.org/W2283866686","https://openalex.org/W4287182096","https://openalex.org/W27867058"],"abstract_inverted_index":{"General":[0],"Matrix":[1],"Multiplication":[2],"(GEMM)":[3],"is":[4,23,108],"a":[5,62,65,103,121,143,156,213,225],"crucial":[6],"algorithm":[7],"for":[8,25,35,91,159,163,261],"various":[9],"applications":[10],"such":[11],"as":[12],"machine":[13],"learning":[14],"and":[15,54,78,97,100,127,150,166,178,257,263],"scientific":[16],"computing":[17,41,84],"since":[18],"an":[19,71],"efficient":[20],"GEMM":[21,68,92,105,139,168,189,222],"implementation":[22,106],"essential":[24],"the":[26,43,94,114,129,137,197,202,246],"performance":[27,37,194],"of":[28,46,64,146,237,253],"these":[29,47],"calculations.":[30],"While":[31],"researchers":[32],"often":[33],"strive":[34],"faster":[36,112],"by":[38,270],"using":[39],"large":[40],"platforms,":[42],"increased":[44],"scale":[45],"systems":[48],"can":[49],"raise":[50],"concerns":[51],"about":[52],"hardware":[53],"software":[55],"reliability.":[56],"In":[57,218],"this":[58],"paper,":[59],"we":[60,154],"present":[61,120,155],"design":[63],"high-performance":[66],"GPU-based":[67],"that":[69,76,107,186],"integrates":[70],"algorithm-based":[72],"fault":[73,134],"tolerance":[74,135],"scheme":[75],"detects":[77],"corrects":[79],"silent":[80],"data":[81],"corruptions":[82],"at":[83,93],"units":[85],"on-the-fly.":[86],"We":[87,119,170],"explore":[88],"fault-tolerant":[89,165,206,221,262],"designs":[90],"thread,":[95],"warp,":[96],"threadblock":[98],"levels,":[99],"also":[101],"provide":[102],"baseline":[104,188],"competitive":[109],"with":[110,136,201,235],"or":[111,192],"than":[113],"state-of-the-art,":[115],"closed-source":[116,198],"cuBLAS":[117,233,269],"GEMM.":[118],"kernel":[122],"fusion":[123],"strategy":[124,211],"to":[125,133,196,232,272],"overlap":[126],"mitigate":[128],"memory":[130],"latency":[131],"due":[132],"original":[138],"computation.":[140],"To":[141],"support":[142],"wide":[144],"range":[145],"input":[147],"matrix":[148],"shapes":[149],"reduce":[151],"development":[152],"costs,":[153],"template-based":[157],"approach":[158],"automatic":[160],"code":[161,247],"generation":[162],"both":[164],"non-fault-tolerant":[167,264],"implementations.":[169],"evaluate":[171],"our":[172,187,208,220],"work":[173],"on":[174,216,229],"NVIDIA":[175],"Tesla":[176],"T4":[177],"A100":[179],"server":[180],"GPUs.":[181],"Our":[182],"experimental":[183],"results":[184],"demonstrate":[185],"shows":[190],"comparable":[191],"superior":[193],"compared":[195,231],"cuBLAS.":[199],"Compared":[200],"prior":[203],"state-of-the-art":[204],"non-fused":[205],"GEMM,":[207],"optimal":[209],"fused":[210],"achieves":[212],"39.04%":[214],"speedup":[215],"average.":[217],"addition,":[219],"incurs":[223],"only":[224],"minimal":[226],"overhead":[227],"(8.89%":[228],"average)":[230],"even":[234],"hundreds":[236],"errors":[238],"injected":[239],"per":[240],"minute.":[241],"For":[242],"irregularly":[243],"shaped":[244],"inputs,":[245],"generator-generated":[248],"kernels":[249],"show":[250],"remarkable":[251],"speedups":[252],"160%":[254],"~":[255,259],"183.5%":[256],"148.55%":[258],"165.12%":[260],"GEMMs,":[265],"respectively,":[266],"which":[267],"outperforms":[268],"up":[271],"41.40%.":[273]},"counts_by_year":[{"year":2025,"cited_by_count":6},{"year":2024,"cited_by_count":3},{"year":2023,"cited_by_count":2}],"updated_date":"2026-03-27T05:58:40.876381","created_date":"2025-10-10T00:00:00"}