Don't let that CPU sit idle: Hardware-aware heterogeneous general matrix multiplication (GEMM)

dc.contributor.authorWarawa, Johnathan
dc.contributor.authorChester, Sean
dc.date.accessioned2026-04-20T21:43:30Z
dc.date.available2026-04-20T21:43:30Z
dc.date.issued2026
dc.description.abstractMatrix multiplication is a fundamental operation used to train neural networks for machine learning. GPUs are well-optimized for several stages of this operation and are thus used to accelerate the work, however, GPUs must be "hosted" by CPUs that remain underutilized while the GPU works, burning cycles that could be put to use by a more sophisticated, heterogeneous algorithm that makes use of both the GPU and CPU at the same time. In this project, we increase the speed of these operations beyond what a single processor could accomplish by developing a heterogeneous algorithm which efficiently divides and interleaves these operations.
dc.description.reviewstatusReviewed
dc.description.scholarlevelUndergraduate
dc.description.sponsorshipJamie Cassels Undergraduate Research Awards (JCURA)
dc.identifier.urihttps://hdl.handle.net/1828/23644
dc.language.isoen
dc.publisherUniversity of Victoria
dc.subjectGPU
dc.subjectSIMD
dc.subjectheterogeneous computing
dc.subjectGEMM
dc.subjectlinear algebra
dc.subjecttensor cores
dc.subjectJamie Cassels Undergraduate Research Awards (JCURA)
dc.subject.departmentDepartment of Computer Science
dc.titleDon't let that CPU sit idle: Hardware-aware heterogeneous general matrix multiplication (GEMM)
dc.typePoster

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
warawa_johnathan_jcura_poster_2026.pdf
Size:
226.22 KB
Format:
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.62 KB
Format:
Item-specific license agreed upon to submission
Description: