Skip to content

Conversation

@lifflander
Copy link
Contributor

@lifflander lifflander commented Dec 18, 2024

Fixes #1

I'm thinking about adding MPI_Get_processor_name in this or we could do it in a script instead.

@lifflander lifflander linked an issue Dec 18, 2024 that may be closed by this pull request
@lifflander
Copy link
Contributor Author

lifflander commented Dec 18, 2024

This is the output it will produce on 8 ranks and 20 iterations that is meant to be read in by a script:

gather: 0: 3.26723: breakdown: 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272 0.154272
gather: 1: 3.21715: breakdown: 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019 0.149019
gather: 2: 3.24052: breakdown: 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853 0.151853
gather: 3: 3.28802: breakdown: 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015 0.149015
gather: 4: 3.27613: breakdown: 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472 0.147472
gather: 5: 3.27223: breakdown: 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965 0.152965
gather: 6: 3.29078: breakdown: 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535 0.147535
gather: 7: 3.25363: breakdown: 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264 0.151264

@cwschilly cwschilly changed the title #1: initial: implement slow node benchmark #1: Implement new benchmark Dec 18, 2024
)

target_link_libraries(${SLOW_NODE_EXE} PUBLIC MPI::MPI_CXX)
target_link_libraries(${SLOW_NODE_EXE} PUBLIC Kokkos::kokkoskernels)
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I had to change this line to link against my installed Trilinos:

target_link_libraries(${SLOW_NODE_EXE} PUBLIC KokkosKernels::kokkoskernels)

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, we need an option to switch between finding Trilinos and linking that way vs. directly using Kokkos kernels.

@lifflander lifflander merged commit 1176e51 into master Mar 4, 2025
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Implement new benchmark

4 participants