{"id":"https://openalex.org/W4406172046","doi":"https://doi.org/10.1109/scw63240.2024.00028","title":"Impacts of floating-point non-associativity on reproducibility for HPC and deep learning applications","display_name":"Impacts of floating-point non-associativity on reproducibility for HPC and deep learning applications","publication_year":2024,"publication_date":"2024-11-17","ids":{"openalex":"https://openalex.org/W4406172046","doi":"https://doi.org/10.1109/scw63240.2024.00028"},"language":"en","primary_location":{"id":"doi:10.1109/scw63240.2024.00028","is_oa":false,"landing_page_url":"https://doi.org/10.1109/scw63240.2024.00028","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","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/A5092687426","display_name":"Sanjif Shanmugavelu","orcid":null},"institutions":[{"id":"https://openalex.org/I4210143063","display_name":"Maxeler Technologies (United Kingdom)","ror":"https://ror.org/05bm26z17","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210143063"]}],"countries":["GB"],"is_corresponding":true,"raw_author_name":"Sanjif Shanmugavelu","raw_affiliation_strings":["Maxeler Technologies, a Groq Company,London,UK,W6 0ND"],"affiliations":[{"raw_affiliation_string":"Maxeler Technologies, a Groq Company,London,UK,W6 0ND","institution_ids":["https://openalex.org/I4210143063"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5009480253","display_name":"Mathieu Taillefumier","orcid":"https://orcid.org/0000-0002-3983-5625"},"institutions":[{"id":"https://openalex.org/I59105498","display_name":"Swisscom (Switzerland)","ror":"https://ror.org/04t1f4f50","country_code":"CH","type":"company","lineage":["https://openalex.org/I59105498"]},{"id":"https://openalex.org/I4391767926","display_name":"CSCS - Swiss National Supercomputing Centre","ror":"https://ror.org/04rzmms09","country_code":null,"type":"facility","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088","https://openalex.org/I4391767926"]},{"id":"https://openalex.org/I35440088","display_name":"ETH Zurich","ror":"https://ror.org/05a28rw58","country_code":"CH","type":"education","lineage":["https://openalex.org/I2799323385","https://openalex.org/I35440088"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Mathieu Taillefumier","raw_affiliation_strings":["ETH Zurich/Swiss National Supercomputing Centre (CSCS),Lugano,Switzerland,6900"],"affiliations":[{"raw_affiliation_string":"ETH Zurich/Swiss National Supercomputing Centre (CSCS),Lugano,Switzerland,6900","institution_ids":["https://openalex.org/I59105498","https://openalex.org/I35440088","https://openalex.org/I4391767926"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5099111547","display_name":"Christopher Culver","orcid":null},"institutions":[{"id":"https://openalex.org/I4210143063","display_name":"Maxeler Technologies (United Kingdom)","ror":"https://ror.org/05bm26z17","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210143063"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Christopher Culver","raw_affiliation_strings":["Maxeler Technologies, a Groq Company,London,UK,W6 0ND"],"affiliations":[{"raw_affiliation_string":"Maxeler Technologies, a Groq Company,London,UK,W6 0ND","institution_ids":["https://openalex.org/I4210143063"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5103133997","display_name":"\u00d3scar Hern\u00e1ndez","orcid":"https://orcid.org/0000-0003-4476-0055"},"institutions":[{"id":"https://openalex.org/I1289243028","display_name":"Oak Ridge National Laboratory","ror":"https://ror.org/01qz5mb56","country_code":"US","type":"facility","lineage":["https://openalex.org/I1289243028","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I4210159294"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Oscar Hernandez","raw_affiliation_strings":["Oak Ridge National Laboratory,Oak Ridge,TN,USA"],"affiliations":[{"raw_affiliation_string":"Oak Ridge National Laboratory,Oak Ridge,TN,USA","institution_ids":["https://openalex.org/I1289243028"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068431797","display_name":"Mark Coletti","orcid":"https://orcid.org/0000-0003-1020-531X"},"institutions":[{"id":"https://openalex.org/I1289243028","display_name":"Oak Ridge National Laboratory","ror":"https://ror.org/01qz5mb56","country_code":"US","type":"facility","lineage":["https://openalex.org/I1289243028","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I4210159294"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Mark Coletti","raw_affiliation_strings":["Oak Ridge National Laboratory,Oak Ridge,TN,USA"],"affiliations":[{"raw_affiliation_string":"Oak Ridge National Laboratory,Oak Ridge,TN,USA","institution_ids":["https://openalex.org/I1289243028"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5060608335","display_name":"Ada Sedova","orcid":"https://orcid.org/0000-0002-8233-3057"},"institutions":[{"id":"https://openalex.org/I1289243028","display_name":"Oak Ridge National Laboratory","ror":"https://ror.org/01qz5mb56","country_code":"US","type":"facility","lineage":["https://openalex.org/I1289243028","https://openalex.org/I1330989302","https://openalex.org/I39565521","https://openalex.org/I4210159294"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Ada Sedova","raw_affiliation_strings":["Oak Ridge National Laboratory,Oak Ridge,TN,USA"],"affiliations":[{"raw_affiliation_string":"Oak Ridge National Laboratory,Oak Ridge,TN,USA","institution_ids":["https://openalex.org/I1289243028"]}]}],"institutions":[],"countries_distinct_count":3,"institutions_distinct_count":6,"corresponding_author_ids":["https://openalex.org/A5092687426"],"corresponding_institution_ids":["https://openalex.org/I4210143063"],"apc_list":null,"apc_paid":null,"fwci":1.45,"has_fulltext":false,"cited_by_count":4,"citation_normalized_percentile":{"value":0.84445266,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":96,"max":99},"biblio":{"volume":null,"issue":null,"first_page":"170","last_page":"179"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.9962000250816345,"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"}},"topics":[{"id":"https://openalex.org/T11181","display_name":"Advanced Data Storage Technologies","score":0.9962000250816345,"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/T10054","display_name":"Parallel Computing and Optimization Techniques","score":0.9894000291824341,"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/T10320","display_name":"Neural Networks and Applications","score":0.98089998960495,"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/reproducibility","display_name":"Reproducibility","score":0.7068423628807068},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6500014066696167},{"id":"https://openalex.org/keywords/associative-property","display_name":"Associative property","score":0.49485987424850464},{"id":"https://openalex.org/keywords/point","display_name":"Point (geometry)","score":0.47845983505249023},{"id":"https://openalex.org/keywords/computer-architecture","display_name":"Computer architecture","score":0.3912351131439209},{"id":"https://openalex.org/keywords/parallel-computing","display_name":"Parallel computing","score":0.376995325088501},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.11382085084915161},{"id":"https://openalex.org/keywords/statistics","display_name":"Statistics","score":0.07539120316505432}],"concepts":[{"id":"https://openalex.org/C9893847","wikidata":"https://www.wikidata.org/wiki/Q1425625","display_name":"Reproducibility","level":2,"score":0.7068423628807068},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6500014066696167},{"id":"https://openalex.org/C159423971","wikidata":"https://www.wikidata.org/wiki/Q177251","display_name":"Associative property","level":2,"score":0.49485987424850464},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.47845983505249023},{"id":"https://openalex.org/C118524514","wikidata":"https://www.wikidata.org/wiki/Q173212","display_name":"Computer architecture","level":1,"score":0.3912351131439209},{"id":"https://openalex.org/C173608175","wikidata":"https://www.wikidata.org/wiki/Q232661","display_name":"Parallel computing","level":1,"score":0.376995325088501},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.11382085084915161},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.07539120316505432},{"id":"https://openalex.org/C2524010","wikidata":"https://www.wikidata.org/wiki/Q8087","display_name":"Geometry","level":1,"score":0.0},{"id":"https://openalex.org/C202444582","wikidata":"https://www.wikidata.org/wiki/Q837863","display_name":"Pure mathematics","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1109/scw63240.2024.00028","is_oa":false,"landing_page_url":"https://doi.org/10.1109/scw63240.2024.00028","pdf_url":null,"source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"SC24-W: Workshops of the International Conference for High Performance Computing, Networking, Storage and Analysis","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320332359","display_name":"Office of Science","ror":"https://ror.org/00mmn6b08"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":24,"referenced_works":["https://openalex.org/W1534870518","https://openalex.org/W2020804487","https://openalex.org/W2103325328","https://openalex.org/W2775708988","https://openalex.org/W2945097505","https://openalex.org/W3027308092","https://openalex.org/W3041643290","https://openalex.org/W3042416028","https://openalex.org/W3088735455","https://openalex.org/W3132277775","https://openalex.org/W3154326799","https://openalex.org/W4225405705","https://openalex.org/W4280589740","https://openalex.org/W4281708879","https://openalex.org/W4285815181","https://openalex.org/W4288079730","https://openalex.org/W4320062103","https://openalex.org/W4327594616","https://openalex.org/W4385406529","https://openalex.org/W4386124351","https://openalex.org/W4389141459","https://openalex.org/W6684754296","https://openalex.org/W6775039030","https://openalex.org/W6790150266"],"related_works":["https://openalex.org/W2413717610","https://openalex.org/W1973270181","https://openalex.org/W2417696084","https://openalex.org/W2368782778","https://openalex.org/W2087830269","https://openalex.org/W3106281778","https://openalex.org/W4324138256","https://openalex.org/W3134060498","https://openalex.org/W2003643616","https://openalex.org/W638612878"],"abstract_inverted_index":{"Run":[0],"to":[1,14,22,49,55,90,157],"run":[2,156,158],"variability":[3,159],"in":[4,18,77,136],"parallel":[5,112],"programs":[6],"caused":[7],"by":[8,183],"floating-point":[9,50,108],"non-associativity":[10,51,109],"has":[11,52],"been":[12,53],"known":[13],"significantly":[15],"affect":[16,28],"reproducibility":[17,212],"iterative":[19],"algorithms,":[20],"due":[21],"accumulating":[23],"errors.":[24],"Non-reproducibility":[25],"can":[26,60,209],"critically":[27],"the":[29,40,104,132,139,150,163,168,173,187,197,203],"efficiency":[30],"and":[31,46,70,72,95,115,118,148,160,165,213],"effectiveness":[32],"of":[33,42,68,93,103,107,121,141,152,167,175,196],"correctness":[34,214],"testing":[35,96],"for":[36,63,144,193],"stochastic":[37],"programs.":[38],"Recently,":[39],"sensitivity":[41],"deep":[43,83,145,198],"learning":[44,84,199],"training":[45],"inference":[47,194],"pipelines":[48],"found":[54],"sometimes":[56],"be":[57,181],"extreme.":[58],"It":[59],"prevent":[61],"certification":[62],"commercial":[64],"applications,":[65],"accurate":[66],"assessment":[67],"robustness":[69],"sensitivity,":[71],"bug":[73],"detection.":[74],"New":[75],"approaches":[76],"scientific":[78],"computing":[79],"applications":[80],"have":[81],"coupled":[82],"models":[85],"with":[86,125],"high-performance":[87],"computing,":[88],"leading":[89],"an":[91,101],"aggravation":[92],"debugging":[94],"challenges.":[97],"Here":[98],"we":[99,171],"perform":[100],"investigation":[102],"statistical":[105],"properties":[106],"within":[110,138,211],"modern":[111],"programming":[113],"models,":[114],"analyze":[116],"performance":[117],"productivity":[119],"impacts":[120,151],"replacing":[122],"atomic":[123],"operations":[124],"deterministic":[126,134,184],"alternatives":[127],"on":[128,162],"GPUs.":[129],"We":[130,201],"examine":[131],"recently-added":[133],"options":[135],"PyTorch":[137],"context":[140],"GPU":[142],"deployment":[143],"learning,":[146],"uncovering":[147],"quantifying":[149],"input":[153],"parameters":[154],"triggering":[155],"reporting":[161],"reliability":[164],"completeness":[166],"documentation.":[169],"Finally,":[170],"evaluate":[172],"strategy":[174,208],"exploiting":[176],"automatic":[177],"determinism":[178],"that":[179,205],"could":[180],"provided":[182],"hardware,":[185],"using":[186],"Groq":[188],"LPU<sup":[189],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[190],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">TM</sup>":[191],"accelerator":[192],"portions":[195],"pipeline.":[200],"demonstrate":[202],"benefits":[204],"a":[206],"hardware-based":[207],"provide":[210],"efforts.":[215]},"counts_by_year":[{"year":2026,"cited_by_count":1},{"year":2025,"cited_by_count":3}],"updated_date":"2026-03-13T16:22:10.518609","created_date":"2025-10-10T00:00:00"}