{"id":{"repo_id":"ubc","oai_identifier":"oai:circle.library.ubc.ca:2429/44729"},"canonical_url":"https://search.ndltd.org/etd/ubc/oai:circle.library.ubc.ca:2429/44729","repository":{"repo_id":"ubc","name":"University of British Columbia","base_url":"http://circle.library.ubc.ca/oai/request"},"display":{"title":"Deterministic execution on GPU architectures","abstract":"Nondeterminism is a key challenge in developing multithreaded applications. Even with the same input, each execution of a multithreaded program may produce a different output. This behavior complicates debugging and limits one’s ability to test for correctness. This non-reproducibility situation is aggravated on massively parallel architectures like graphics processing units (GPUs) with thousands of concurrent threads. We believe providing a deterministic environment to ease debugging and testing of GPU applications is essential to enable a broader class of software to use GPUs. Many hardware and software techniques have been proposed for providing determinism on general-purpose multi-core processors. However, these techniques are designed for small numbers of threads. Scaling them to thousands of threads on a GPU is a major challenge. Here we propose a scalable hardware mechanism, GPUDet, to provide determinism in GPU architectures. In this thesis we characterize the existing deterministic and nondeterministic aspects of current GPU execution models, and we use these observations to inform GPUDet’s design. For example, GPUDet leverages the inherent determinism of the SIMD hardware in GPUs to provide determinism within a wavefront at no cost. GPUDet also exploits the Z-Buffer Unit, an existing GPU hardware unit for graphics rendering, to allow parallel out-of-order memory writes to produce a deterministic output. Other optimizations in GPUDet include deterministic parallel execution of atomic operations and a workgroup-aware algorithm that eliminates unnecessary global synchronizations. Our simulation results indicate that GPUDet incurs only 2× slowdown on average over a baseline nondeterministic architecture, with runtime overheads as low as 4% for compute-bound applications, despite running GPU kernels with thousands of threads. We also characterize the sources of overhead for deterministic execution on GPUs to provide insights for further optimizations.","abstract_html":"Nondeterminism is a key challenge in developing multithreaded applications. Even with the same input, each execution of a multithreaded program may produce a different output. This behavior complicates debugging and limits one’s ability to test for correctness. This non-reproducibility situation is aggravated on massively parallel architectures like graphics processing units (GPUs) with thousands of concurrent threads. We believe providing a deterministic environment to ease debugging and testing of GPU applications is essential to enable a broader class of software to use GPUs. Many hardware and software techniques have been proposed for providing determinism on general-purpose multi-core processors. However, these techniques are designed for small numbers of threads. Scaling them to thousands of threads on a GPU is a major challenge. Here we propose a scalable hardware mechanism, GPUDet, to provide determinism in GPU architectures. In this thesis we characterize the existing deterministic and nondeterministic aspects of current GPU execution models, and we use these observations to inform GPUDet’s design. For example, GPUDet leverages the inherent determinism of the SIMD hardware in GPUs to provide determinism within a wavefront at no cost. GPUDet also exploits the Z-Buffer Unit, an existing GPU hardware unit for graphics rendering, to allow parallel out-of-order memory writes to produce a deterministic output. Other optimizations in GPUDet include deterministic parallel execution of atomic operations and a workgroup-aware algorithm that eliminates unnecessary global synchronizations. Our simulation results indicate that GPUDet incurs only 2× slowdown on average over a baseline nondeterministic architecture, with runtime overheads as low as 4% for compute-bound applications, despite running GPU kernels with thousands of threads. We also characterize the sources of overhead for deterministic execution on GPUs to provide insights for further optimizations.","abstract_has_math":false,"creators":["Jooybar, Mohammad Hadi"],"institution":"University of British Columbia","degree_name":"Master of Applied Science - MASc","degree_level":"master's","degree_discipline":"Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2013,"date_issued":"2013","date_published":"2013","updated_at":"2026-11-19T14:57:44Z","subjects":[],"languages":["eng"],"rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"rights_urls":["http://creativecommons.org/licenses/by-nc-nd/4.0/"],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2429/44729","outbound_label":"Handle","outbound_source":"dc:identifier","record_url":"http://hdl.handle.net/2429/44729","record_label":"Repository record","record_kind":"handle","record_source":"dc:identifier","download_url":"http://circle.library.ubc.ca/bitstream/2429/44729/1/ubc_2013_fall_jooybar_mohammad+hadi.pdf","download_label":"Download PDF","download_source":"dc:identifier","access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"http://circle.library.ubc.ca/oai/request?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Acircle.library.ubc.ca%3A2429%2F44729","prefix":"etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Jooybar, Mohammad Hadi"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2013"]},{"key":"dc:publisher","label":"Institution","values":["University of British Columbia"]},{"key":"dc:type","label":"Dc Type","values":["Text"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical and Computer Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["master's"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Applied Science - MASc"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of British Columbia"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["eng"]},{"key":"dc:rights","label":"Dc Rights","values":["http://creativecommons.org/licenses/by-nc-nd/4.0/","Attribution-NonCommercial-NoDerivatives 4.0 International"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2429/44729","http://circle.library.ubc.ca/bitstream/2429/44729/1/ubc_2013_fall_jooybar_mohammad+hadi.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Nondeterminism is a key challenge in developing multithreaded applications. Even with the same input, each execution of a multithreaded program may produce a different output. This behavior complicates debugging and limits one’s ability to test for correctness. This non-reproducibility situation is aggravated on massively parallel architectures like graphics processing units (GPUs) with thousands of concurrent threads. We believe providing a deterministic environment to ease debugging and testing of GPU applications is essential to enable a broader class of software to use GPUs. Many hardware and software techniques have been proposed for providing determinism on general-purpose multi-core processors. However, these techniques are designed for small numbers of threads. Scaling them to thousands of threads on a GPU is a major challenge. Here we propose a scalable hardware mechanism, GPUDet, to provide determinism in GPU architectures. In this thesis we characterize the existing deterministic and nondeterministic aspects of current GPU execution models, and we use these observations to inform GPUDet’s design. For example, GPUDet leverages the inherent determinism of the SIMD hardware in GPUs to provide determinism within a wavefront at no cost. GPUDet also exploits the Z-Buffer Unit, an existing GPU hardware unit for graphics rendering, to allow parallel out-of-order memory writes to produce a deterministic output. Other optimizations in GPUDet include deterministic parallel execution of atomic operations and a workgroup-aware algorithm that eliminates unnecessary global synchronizations. Our simulation results indicate that GPUDet incurs only 2× slowdown on average over a baseline nondeterministic architecture, with runtime overheads as low as 4% for compute-bound applications, despite running GPU kernels with thousands of threads. We also characterize the sources of overhead for deterministic execution on GPUs to provide insights for further optimizations."]},{"key":"dc:format","label":"Dc Format","values":["1180462","application/pdf"]},{"key":"dc:title","label":"Title","values":["Deterministic execution on GPU architectures"]}]}],"canonical_facts":{"dc:creator":["Jooybar, Mohammad Hadi"],"dc:date":["2013"],"dc:description":["Nondeterminism is a key challenge in developing multithreaded applications. Even with the same input, each execution of a multithreaded program may produce a different output. This behavior complicates debugging and limits one’s ability to test for correctness. This non-reproducibility situation is aggravated on massively parallel architectures like graphics processing units (GPUs) with thousands of concurrent threads. We believe providing a deterministic environment to ease debugging and testing of GPU applications is essential to enable a broader class of software to use GPUs. Many hardware and software techniques have been proposed for providing determinism on general-purpose multi-core processors. However, these techniques are designed for small numbers of threads. Scaling them to thousands of threads on a GPU is a major challenge. Here we propose a scalable hardware mechanism, GPUDet, to provide determinism in GPU architectures. In this thesis we characterize the existing deterministic and nondeterministic aspects of current GPU execution models, and we use these observations to inform GPUDet’s design. For example, GPUDet leverages the inherent determinism of the SIMD hardware in GPUs to provide determinism within a wavefront at no cost. GPUDet also exploits the Z-Buffer Unit, an existing GPU hardware unit for graphics rendering, to allow parallel out-of-order memory writes to produce a deterministic output. Other optimizations in GPUDet include deterministic parallel execution of atomic operations and a workgroup-aware algorithm that eliminates unnecessary global synchronizations. Our simulation results indicate that GPUDet incurs only 2× slowdown on average over a baseline nondeterministic architecture, with runtime overheads as low as 4% for compute-bound applications, despite running GPU kernels with thousands of threads. We also characterize the sources of overhead for deterministic execution on GPUs to provide insights for further optimizations."],"dc:format":["1180462","application/pdf"],"dc:identifier":["http://hdl.handle.net/2429/44729","http://circle.library.ubc.ca/bitstream/2429/44729/1/ubc_2013_fall_jooybar_mohammad+hadi.pdf"],"dc:language":["eng"],"dc:publisher":["University of British Columbia"],"dc:rights":["http://creativecommons.org/licenses/by-nc-nd/4.0/","Attribution-NonCommercial-NoDerivatives 4.0 International"],"dc:title":["Deterministic execution on GPU architectures"],"dc:type":["Text"],"thesis:degree_discipline":["Electrical and Computer Engineering"],"thesis:degree_level":["master's"],"thesis:degree_name":["Master of Applied Science - MASc"],"thesis:institution_name":["University of British Columbia"]},"updated_at":"2026-11-19T14:57:44Z"}