Ginkgo's performance on Intel Gen12 hardware
<p>These two performance plots show the performance of the Ginkgo library on Intel Gen12 hardware relative to the peak performance of the hardware. The performance is after an initial porting of the library and some tuning on non-HPC hardware, thus it is expected that the performance is not optimal. We compare our library's performance to the native vendor-provided oneMKL functionality for the Sparse Matrix-Vector product. We see that Ginkgo is at least competitive if not better than the vendor implementation. The solver results show that Ginkgo's solvers are also suitable for Intel hardware although some solvers, in particular GMRES, could be better optimized.</p>
ShareScore
40/100
Overall dataset sharing score
Score breakdown
These five areas show where the dataset supports — or may limit — practical reuse.
- Stewardship
- 8
- Harmonization
- 4
- Access
- 20
- Reuse readiness
- 8
- Engagement
- 0