DC Field | Value | Language |
---|---|---|
dc.contributor.author | Kim, Hyeseong | ko |
dc.contributor.author | Lee, Yunjae | ko |
dc.contributor.author | Rhu, Minsoo | ko |
dc.date.accessioned | 2024-09-03T10:00:13Z | - |
dc.date.available | 2024-09-03T10:00:13Z | - |
dc.date.created | 2024-08-29 | - |
dc.date.issued | 2024-01 | - |
dc.identifier.citation | IEEE COMPUTER ARCHITECTURE LETTERS, v.23, no.1, pp.9 - 10 | - |
dc.identifier.issn | 1556-6056 | - |
dc.identifier.uri | http://hdl.handle.net/10203/322585 | - |
dc.description.abstract | Deep neural network (DNN)-based recommendation systems (RecSys) are one of the most successfully deployed machine learning applications in commercial services for predicting ad click-through rates or rankings. While numerous prior work explored hardware and software solutions to reduce the training time of RecSys, its end-to-end training pipeline including the data preprocessing stage has received little attention. In this work, we provide a comprehensive analysis of RecSys data preprocessing, root-causing the feature generation and normalization steps to cause a major performance bottleneck. Based on our characterization, we explore the efficacy of an FPGA-accelerated RecSys preprocessing system that achieves a significant 3.4-12.1x end-to-end speedup compared to the baseline CPU-based RecSys preprocessing system. | - |
dc.language | English | - |
dc.publisher | IEEE COMPUTER SOC | - |
dc.title | FPGA-Accelerated Data Preprocessing for Personalized Recommendation Systems | - |
dc.type | Article | - |
dc.identifier.wosid | 001165793400001 | - |
dc.identifier.scopusid | 2-s2.0-85179040384 | - |
dc.type.rims | ART | - |
dc.citation.volume | 23 | - |
dc.citation.issue | 1 | - |
dc.citation.beginningpage | 9 | - |
dc.citation.endingpage | 10 | - |
dc.citation.publicationname | IEEE COMPUTER ARCHITECTURE LETTERS | - |
dc.identifier.doi | 10.1109/LCA.2023.3336841 | - |
dc.contributor.localauthor | Rhu, Minsoo | - |
dc.contributor.nonIdAuthor | Kim, Hyeseong | - |
dc.description.isOpenAccess | N | - |
dc.type.journalArticle | Article | - |
dc.subject.keywordAuthor | Personalized recommendation system | - |
dc.subject.keywordAuthor | data preprocessing | - |
dc.subject.keywordAuthor | training | - |
dc.subject.keywordAuthor | FPGA | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.