Skip to content

Conversation

@edopao
Copy link
Contributor

@edopao edopao commented Jan 21, 2026

One innocent line that makes a huge performance difference for muphys graupel program in ICON4Py.

@edopao
Copy link
Contributor Author

edopao commented Jan 21, 2026

bench_blueline_stencil_compute

@edopao edopao changed the title feat[nex-dace]: Enable tasklet fusion in dataflow optimization feat[next-dace]: Enable tasklet fusion in dataflow optimization Jan 21, 2026
@philip-paul-mueller
Copy link
Contributor

There are big changes also in the dycore, it might be only fluctuations, so I would advise to run it again to see what is fluctuations and what is not.

Copy link
Contributor

@philip-paul-mueller philip-paul-mueller left a comment

Choose a reason for hiding this comment

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

See comment.

@edopao edopao force-pushed the dace_tasklet_fusion branch from e58a55d to 19172fd Compare January 21, 2026 16:24
Comment on lines +684 to +686
sdfg.apply_transformations_repeated(
dace_dataflow.TaskletFusion, validate=False, validate_all=validate_all
)
Copy link
Contributor

Choose a reason for hiding this comment

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

The location is not optimial because it essentially destroy LoopBlooking but since that stencil does not depend on it, it should be fine.

@edopao
Copy link
Contributor Author

edopao commented Jan 23, 2026

Surpassed by development in #2457

@edopao edopao closed this Jan 23, 2026
@edopao edopao deleted the dace_tasklet_fusion branch January 23, 2026 12:00
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.

2 participants