{"id":{"repo_id":"cornell","oai_identifier":"oai:ecommons.cornell.edu:1813/7972"},"canonical_url":"https://search.ndltd.org/etd/cornell/oai:ecommons.cornell.edu:1813/7972","repository":{"repo_id":"cornell","name":"Cornell University","base_url":"https://ecommons.cornell.edu/server/oai/request"},"display":{"title":"PREDICTING PARALLEL APPLICATION PERFORMANCE VIA MACHINE LEARNING APPROACHES","abstract":"Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes.","abstract_html":"Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes.","abstract_has_math":false,"creators":["Singh, Karan"],"institution":null,"degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2007,"date_issued":"2007-07-31T15:25:21Z","date_published":"2007-07-31T15:25:21Z","updated_at":"2026-11-19T03:21:11Z","subjects":["high-performance computing","performance modeling","artificial neural networks"],"languages":["en_US"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier.bibid","label":"Dc Identifier Bibid","values":["6476387"],"render_values":[{"text":"6476387","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1813/7972","outbound_label":"Handle","outbound_source":"dc:identifier.uri","record_url":"https://hdl.handle.net/1813/7972","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier.uri","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://ecommons.cornell.edu/server/oai/request?verb=GetRecord&metadataPrefix=dim&identifier=oai%3Aecommons.cornell.edu%3A1813%2F7972","prefix":"dim"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Singh, Karan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2007-07-31T15:25:21Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2012-07-31T06:11:33Z"]},{"key":"dc:date.issued","label":"Date","values":["2007-07-31T15:25:21Z"]},{"key":"dc:type","label":"Dc Type","values":["dissertation or thesis"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["high-performance computing","performance modeling","artificial neural networks"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en_US"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.bibid","label":"Dc Identifier Bibid","values":["6476387"]},{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1813/7972"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"cerif:openaire","label":"Cerif Openaire","values":["<Publication xmlns=\"https://www.openaire.eu/cerif-profile/1.1/\" id=\"a92a7bb3-e52d-4cd3-8e5d-323817ce72f5\"> <Type xmlns=\"https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types\">http://purl.org/coar/resource_type/c_1843</Type> <Language>en_US</Language> <Title>PREDICTING PARALLEL APPLICATION PERFORMANCE VIA MACHINE LEARNING APPROACHES</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2007-07-31T15:25:21Z</PublicationDate> <Authors> <Author> <DisplayName>Singh, Karan</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <OrgUnit /> </Publisher> </Publishers> <Keyword>high-performance computing</Keyword> <Keyword>performance modeling</Keyword> <Keyword>artificial neural networks</Keyword> <Abstract>Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes.</Abstract> <Access xmlns=\"http://purl.org/coar/access_right\" > </Access> </Publication>"]},{"key":"cris:virtual.author","label":"Cris Virtual Author","values":["Singh, Karan"]},{"key":"cris:virtualsource.collection","label":"Cris Virtualsource Collection","values":["5893a6ea-7af3-41d7-abc6-04bcd26ab5df"]},{"key":"dc:description.abstract","label":"Abstract","values":["Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes."]},{"key":"dc:description.collection","label":"Dc Description Collection","values":["Cornell Theses and Dissertations"]},{"key":"dc:description.sponsorship","label":"Dc Description Sponsorship","values":["National Science Foundation Grant Number CCF-0444413; United States Department of Energy Grant Number W-7405-Eng-48"]},{"key":"dc:format.extent","label":"Dc Format Extent","values":["296834 bytes"]},{"key":"dc:format.mimetype","label":"Dc Format Mimetype","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["PREDICTING PARALLEL APPLICATION PERFORMANCE VIA MACHINE LEARNING APPROACHES"]},{"key":"dspace:entity.type","label":"Dspace Entity Type","values":["Publication"]},{"key":"others:access_status","label":"Others Access Status","values":["open.access"]}]}],"canonical_facts":{"cerif:openaire":["<Publication xmlns=\"https://www.openaire.eu/cerif-profile/1.1/\" id=\"a92a7bb3-e52d-4cd3-8e5d-323817ce72f5\"> <Type xmlns=\"https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types\">http://purl.org/coar/resource_type/c_1843</Type> <Language>en_US</Language> <Title>PREDICTING PARALLEL APPLICATION PERFORMANCE VIA MACHINE LEARNING APPROACHES</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2007-07-31T15:25:21Z</PublicationDate> <Authors> <Author> <DisplayName>Singh, Karan</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <OrgUnit /> </Publisher> </Publishers> <Keyword>high-performance computing</Keyword> <Keyword>performance modeling</Keyword> <Keyword>artificial neural networks</Keyword> <Abstract>Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes.</Abstract> <Access xmlns=\"http://purl.org/coar/access_right\" > </Access> </Publication>"],"cris:virtual.author":["Singh, Karan"],"cris:virtualsource.collection":["5893a6ea-7af3-41d7-abc6-04bcd26ab5df"],"dc:creator":["Singh, Karan"],"dc:date.accessioned":["2007-07-31T15:25:21Z"],"dc:date.available":["2012-07-31T06:11:33Z"],"dc:date.issued":["2007-07-31T15:25:21Z"],"dc:description.abstract":["Consistently growing architectural complexity and machine scales make creating accurate performance models for large-scale applications increasingly challenging. Traditional analytic models are difficult and time-consuming to construct, and are often unable to capture full system and application complexity. To address these challenges, we automatically build models based on execution samples. We use multilayer neural networks, since they can represent arbitrary functions and handle noisy inputs robustly. In this thesis, we focus on two well known parallel applications whose variations in execution times are not well understood: SMG2000, a semicoarsening multigrid solver, and HPL, an open source implementation of LINPACK. We sparsely sample performance data on two radically different platforms across large, multi-dimensional parameter spaces and show that our models based on this data can predict performance within 2% to 7% of actual application runtimes."],"dc:description.collection":["Cornell Theses and Dissertations"],"dc:description.sponsorship":["National Science Foundation Grant Number CCF-0444413; United States Department of Energy Grant Number W-7405-Eng-48"],"dc:format.extent":["296834 bytes"],"dc:format.mimetype":["application/pdf"],"dc:identifier.bibid":["6476387"],"dc:identifier.uri":["https://hdl.handle.net/1813/7972"],"dc:language.iso":["en_US"],"dc:subject":["high-performance computing","performance modeling","artificial neural networks"],"dc:title":["PREDICTING PARALLEL APPLICATION PERFORMANCE VIA MACHINE LEARNING APPROACHES"],"dc:type":["dissertation or thesis"],"dspace:entity.type":["Publication"],"others:access_status":["open.access"]},"updated_at":"2026-11-19T03:21:11Z"}