DIESEL: A novel deep learning-based tool for SpMV computations and solving sparse linear equation systems

dc.contributorAalto-yliopistofi
dc.contributorAalto Universityen
dc.contributor.authorMohammed, Thahaen_US
dc.contributor.authorAlbeshri, Aiiaden_US
dc.contributor.authorKatib, Iyaden_US
dc.contributor.authorMehmood, Rashiden_US
dc.contributor.departmentDepartment of Computer Scienceen
dc.contributor.groupauthorProfessorship Di Francesco Marioen
dc.contributor.organizationKing Abdulaziz Universityen_US
dc.date.accessioned2021-12-01T07:50:43Z
dc.date.available2021-12-01T07:50:43Z
dc.date.embargoinfo:eu-repo/date/embargoEnd/2021-11-30en_US
dc.date.issued2021-06en_US
dc.description.abstractSparse linear algebra is central to many areas of engineering, science, and business. The community has done considerable work on proposing new methods for sparse matrix-vector multiplication (SpMV) computations and iterative sparse solvers on graphical processing units (GPUs). Due to vast variations in matrix features, no single method performs well across all sparse matrices. A few tools on automatic prediction of best-performing SpMV kernels have emerged recently and require many more efforts to fully utilize their potential. The utilization of a GPU by the existing SpMV kernels is far from its full capacity. Moreover, the development and performance analysis of SpMV techniques on GPUs have not been studied in sufficient depth. This paper proposes DIESEL, a deep learning-based tool that predicts and executes the best performing SpMV kernel for a given matrix using a feature set carefully devised by us through rigorous empirical and mathematical instruments. The dataset comprises 1056 matrices from 26 different real-life application domains including computational fluid dynamics, materials, electromagnetics, economics, and more. We propose a range of new metrics and methods for performance analysis, visualization, and comparison of SpMV tools. DIESEL provides better performance with its accuracy 88.2%, workload accuracy 91.96%, and average relative loss 4.4%, compared to 85.9%, 85.31%, and 7.65% by the next best performing artificial intelligence (AI)-based SpMV tool. The extensive results and analyses presented in this paper provide several key insights into the performance of the SpMV tools and how these relate to the matrix datasets and the performance metrics, allowing the community to further improve and compare basic and AI-based SpMV tools in the future.en
dc.description.versionPeer revieweden
dc.format.extent43
dc.format.extent6313–6355
dc.format.mimetypeapplication/pdfen_US
dc.identifier.citationMohammed, T, Albeshri, A, Katib, I & Mehmood, R 2021, ' DIESEL: A novel deep learning-based tool for SpMV computations and solving sparse linear equation systems ', JOURNAL OF SUPERCOMPUTING, vol. 77, no. 6, pp. 6313–6355 . https://doi.org/10.1007/s11227-020-03489-3en
dc.identifier.doi10.1007/s11227-020-03489-3en_US
dc.identifier.issn0920-8542
dc.identifier.otherPURE UUID: 447f5121-eb67-4c52-ab7b-86c6640a2d26en_US
dc.identifier.otherPURE ITEMURL: https://research.aalto.fi/en/publications/447f5121-eb67-4c52-ab7b-86c6640a2d26en_US
dc.identifier.otherPURE LINK: http://www.scopus.com/inward/record.url?scp=85097009275&partnerID=8YFLogxKen_US
dc.identifier.otherPURE FILEURL: https://research.aalto.fi/files/76178996/SCI_Thaha_etal_DIESEL_A_Novel_Deep_Learning_Based_Tool_for_SPMV_Journal_of_Supercomputing_2021.pdfen_US
dc.identifier.urihttps://aaltodoc.aalto.fi/handle/123456789/111347
dc.identifier.urnURN:NBN:fi:aalto-2021120110497
dc.language.isoenen
dc.publisherSpringer Netherlands
dc.relation.ispartofseriesJOURNAL OF SUPERCOMPUTINGen
dc.relation.ispartofseriesVolume 77, issue 6en
dc.rightsopenAccessen
dc.subject.keywordSparse linear algebraen_US
dc.subject.keywordSparse linear equation systemsen_US
dc.subject.keywordSparse matrix vector product (SpMV)en_US
dc.subject.keywordGraphics processing units (GPUs)en_US
dc.subject.keywordDeep Learningen_US
dc.subject.keywordartificial intelligence (AI)en_US
dc.subject.keywordIterative solversen_US
dc.titleDIESEL: A novel deep learning-based tool for SpMV computations and solving sparse linear equation systemsen
dc.typeA1 Alkuperäisartikkeli tieteellisessä aikakauslehdessäfi
dc.type.versionacceptedVersion
Files