Joint Schedule and Layout Autotuning for Sparse Matrices with Compound Entries on GPUs

dc.contributor.authorMueller-Roemer, Johannes Sebastianen_US
dc.contributor.authorStork, Andréen_US
dc.contributor.authorFellner, Dieter W.en_US
dc.contributor.editorSchulz, Hans-Jörg and Teschner, Matthias and Wimmer, Michaelen_US
dc.date.accessioned2019-09-29T06:46:05Z
dc.date.available2019-09-29T06:46:05Z
dc.date.issued2019
dc.description.abstractLarge sparse matrices with compound entries, i.e., complex and quaternionic matrices as well as matrices with dense blocks, are a core component of many algorithms in geometry processing, physically based animation, and other areas of computer graphics. We generalize several matrix layouts and apply joint schedule and layout autotuning to improve the performance of the sparse matrix-vector product on massively parallel graphics processing units. Compared to schedule tuning without layout tuning, we achieve speedups of up to 5.5x. In comparison to cuSPARSE, we achieve speedups of up to 4.7xen_US
dc.description.sectionheadersGPU
dc.description.seriesinformationVision, Modeling and Visualization
dc.identifier.doi10.2312/vmv.20191324
dc.identifier.isbn978-3-03868-098-7
dc.identifier.pages109-116
dc.identifier.urihttps://doi.org/10.2312/vmv.20191324
dc.identifier.urihttps://diglib.eg.org:443/handle/10.2312/vmv20191324
dc.publisherThe Eurographics Associationen_US
dc.subjectComputing methodologies
dc.subjectMassively parallel algorithms
dc.subjectParallel programming languages
dc.subjectMathematics of computing
dc.subjectComputations on matrices
dc.titleJoint Schedule and Layout Autotuning for Sparse Matrices with Compound Entries on GPUsen_US
Files
Original bundle
Now showing 1 - 2 of 2
Loading...
Thumbnail Image
Name:
109-116.pdf
Size:
253.83 KB
Format:
Adobe Portable Document Format
No Thumbnail Available
Name:
supplemental.zip
Size:
2.97 MB
Format:
Zip file
Collections