{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/108171"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/108171","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"An optimizing compiler for ONNX models on heterogeneous systems","abstract":"In order to build, train, and deploy deep learning models for modern data-driven applications, programs need to be executed on top of specialized heterogeneous systems for better performance. However, programming on those heterogeneous systems remains a fundamental challenge in terms of the interoperability issue between high-level deep learning frameworks and the programmability issue between different low-level heterogeneous systems. In this work, we propose a portable and highly optimizing compiler for neural network models, which is based on an open format - ONNX of deep learning models, running on heterogeneous systems. It consists of a front-end and a back-end to address those above issues. The goal of this neural network compiler is also to map high-level neural network models to low-level executable programs. We evaluate this work with several deep learning neural network models and our neural network compiler is able to outperform ONNX runtime by up to 3.15x and Keras by up to 4.37x on certain workloads.","abstract_html":"In order to build, train, and deploy deep learning models for modern data-driven applications, programs need to be executed on top of specialized heterogeneous systems for better performance. However, programming on those heterogeneous systems remains a fundamental challenge in terms of the interoperability issue between high-level deep learning frameworks and the programmability issue between different low-level heterogeneous systems. In this work, we propose a portable and highly optimizing compiler for neural network models, which is based on an open format - ONNX of deep learning models, running on heterogeneous systems. It consists of a front-end and a back-end to address those above issues. The goal of this neural network compiler is also to map high-level neural network models to low-level executable programs. We evaluate this work with several deep learning neural network models and our neural network compiler is able to outperform ONNX runtime by up to 3.15x and Keras by up to 4.37x on certain workloads.","abstract_has_math":false,"creators":["Shi, Yuanjing"],"institution":"University of Illinois at Urbana-Champaign","degree_name":"M.S.","degree_level":"Thesis","degree_discipline":"Computer Science","degree_department":null,"school":null,"contributors":["Adve, Vikram S."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2020,"date_issued":"2020-08-26T23:58:43Z","date_published":"2020-08-26T23:58:43Z","updated_at":"2026-07-22T22:24:47Z","subjects":["Compiler"],"languages":["en"],"rights":["Copyright 2020 Yuanjing Shi"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"http://hdl.handle.net/2142/108171","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Adve, Vikram S."]},{"key":"dc:creator","label":"Author","values":["Shi, Yuanjing"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2020-08-26T23:58:43Z","2022-08-26T23:58:55Z","2020-05-11","2020-05"]},{"key":"dc:type","label":"Dc Type","values":["text","Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Computer Science"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois at Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Compiler"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2020 Yuanjing Shi"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["http://hdl.handle.net/2142/108171"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["In order to build, train, and deploy deep learning models for modern data-driven applications, programs need to be executed on top of specialized heterogeneous systems for better performance. However, programming on those heterogeneous systems remains a fundamental challenge in terms of the interoperability issue between high-level deep learning frameworks and the programmability issue between different low-level heterogeneous systems. In this work, we propose a portable and highly optimizing compiler for neural network models, which is based on an open format - ONNX of deep learning models, running on heterogeneous systems. It consists of a front-end and a back-end to address those above issues. The goal of this neural network compiler is also to map high-level neural network models to low-level executable programs. We evaluate this work with several deep learning neural network models and our neural network compiler is able to outperform ONNX runtime by up to 3.15x and Keras by up to 4.37x on certain workloads.","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2022-05-01","The student, Yuanjing Shi, accepted the attached license on 2020-05-07 at 17:08.","The student, Yuanjing Shi, submitted this Thesis for approval on 2020-05-07 at 17:09.","This Thesis was approved for publication on 2020-05-11 at 11:14.","DSpace SAF Submission Ingestion Package generated from Vireo submission #15281 on 2020-08-25 at 17:30:29","Made available in DSpace on 2020-08-26T23:58:43Z (GMT). No. of bitstreams: 2 SHI-THESIS-2020.pdf: 616960 bytes, checksum: 404f09d20a829694938d2886a1502c9e (MD5) LICENSE.txt: 4209 bytes, checksum: 3c7cdd537fcb7d58794264791e061f8b (MD5) Previous issue date: 2020-05-11","Embargo set by: Seth Robbins for item 115784 Lift date: 2022-08-26T23:58:55Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only"]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["An optimizing compiler for ONNX models on heterogeneous systems"]}]}],"canonical_facts":{"dc:contributor":["Adve, Vikram S."],"dc:creator":["Shi, Yuanjing"],"dc:date":["2020-08-26T23:58:43Z","2022-08-26T23:58:55Z","2020-05-11","2020-05"],"dc:description":["In order to build, train, and deploy deep learning models for modern data-driven applications, programs need to be executed on top of specialized heterogeneous systems for better performance. However, programming on those heterogeneous systems remains a fundamental challenge in terms of the interoperability issue between high-level deep learning frameworks and the programmability issue between different low-level heterogeneous systems. In this work, we propose a portable and highly optimizing compiler for neural network models, which is based on an open format - ONNX of deep learning models, running on heterogeneous systems. It consists of a front-end and a back-end to address those above issues. The goal of this neural network compiler is also to map high-level neural network models to low-level executable programs. We evaluate this work with several deep learning neural network models and our neural network compiler is able to outperform ONNX runtime by up to 3.15x and Keras by up to 4.37x on certain workloads.","Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2022-05-01","The student, Yuanjing Shi, accepted the attached license on 2020-05-07 at 17:08.","The student, Yuanjing Shi, submitted this Thesis for approval on 2020-05-07 at 17:09.","This Thesis was approved for publication on 2020-05-11 at 11:14.","DSpace SAF Submission Ingestion Package generated from Vireo submission #15281 on 2020-08-25 at 17:30:29","Made available in DSpace on 2020-08-26T23:58:43Z (GMT). No. of bitstreams: 2 SHI-THESIS-2020.pdf: 616960 bytes, checksum: 404f09d20a829694938d2886a1502c9e (MD5) LICENSE.txt: 4209 bytes, checksum: 3c7cdd537fcb7d58794264791e061f8b (MD5) Previous issue date: 2020-05-11","Embargo set by: Seth Robbins for item 115784 Lift date: 2022-08-26T23:58:55Z Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD system","U of I Only"],"dc:format":["application/pdf"],"dc:identifier":["http://hdl.handle.net/2142/108171"],"dc:language":["en"],"dc:rights":["Copyright 2020 Yuanjing Shi"],"dc:subject":["Compiler"],"dc:title":["An optimizing compiler for ONNX models on heterogeneous systems"],"dc:type":["text","Thesis"],"thesis:degree_discipline":["Computer Science"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S."],"thesis:institution_name":["University of Illinois at Urbana-Champaign"]},"updated_at":"2026-07-22T22:24:47Z"}