Compare commits

...

6 Commits

Author SHA1 Message Date
Kenneth Moreland
fa3e781acb Merge branch 'bench-force-cuda' into 'master'
Force CUDA device in performance benchmarks

See merge request vtk/vtk-m!3224
2024-06-28 15:13:07 -04:00
Kenneth Moreland
6b9df58824 Force CUDA device in performance benchmarks 2024-06-28 15:12:22 -04:00
Kenneth Moreland
fc570a75a5 Merge branch 'release' into master 2024-06-28 11:10:45 -04:00
Kenneth Moreland
f610044d79 Merge topic 'split-contour-bench-2-1' into release-2.0
6f5f65487 Split the contour benchmark into structured/unstructured

Acked-by: Kitware Robot <kwrobot@kitware.com>
Acked-by: Vicente Bolea <vicente.bolea@kitware.com>
Merge-request: !3239
2024-06-28 11:10:45 -04:00
Kenneth Moreland
cb07d8400c Merge branch 'release-2.0' into release 2024-06-28 11:10:45 -04:00
Kenneth Moreland
6f5f654878 Split the contour benchmark into structured/unstructured
We've been having problems with PerformanceTestBenchContour. In the last
few iteration, the runtime goes way up. We cannot find any reason for
this in the source code. There don't appear to be any particular
problems with memory or tables. The best we can figure is an issue with
the device hardware in the container.

The easy solution should be to break the benchmark into smaller peices
to avoid the problem.
2024-06-26 17:47:27 -04:00

@ -261,6 +261,7 @@ test:ubuntu1804_cuda_perftest:
variables:
TEST_INCLUSIONS: "PerformanceTest"
VTKm_PERF_REMOTE_URL: "https://vbolea:$VTKM_BENCH_RECORDS_TOKEN@gitlab.kitware.com/vbolea/vtk-m-benchmark-records.git"
VTKm_PERF_BENCH_DEVICE: "cuda"
VTKm_PERF_ALPHA: "0.05"
VTKm_PERF_REPETITIONS: "10"
VTKm_PERF_DIST: "t"