Skip to content

Adjust num_chunks after fixing bug in PR #152.#155

Merged
henricasanova merged 6 commits intomainfrom
num_chunks
May 9, 2026
Merged

Adjust num_chunks after fixing bug in PR #152.#155
henricasanova merged 6 commits intomainfrom
num_chunks

Conversation

@quantumsteve
Copy link
Copy Markdown
Collaborator

3000000 / 16384 = 183, or approximately 200.

Signed-off-by: Steven Hahn <hahnse@ornl.gov>
@quantumsteve quantumsteve requested a review from henricasanova May 8, 2026 15:20
@henricasanova
Copy link
Copy Markdown
Contributor

Let me look into this. I just ran wfbench on my laptop, and it's not really behaving as expected. For instance, if I do a 10s time limit it runs in almost zero time. I'll investigate....

@henricasanova
Copy link
Copy Markdown
Contributor

I pushed a few fixes/changes based on things I noticed. Regarding the min_chunk_size_cpu_work = X * min_chunk_size_time line, using X=2500 produces the "1 second on Henri's laptop" behavior. It's perhaps a weird reference point, but the point it to avoid having an inordinate number of chunks.

Copy link
Copy Markdown
Contributor

@henricasanova henricasanova left a comment

Choose a reason for hiding this comment

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

I think this is all good to merge in.

@henricasanova henricasanova merged commit dd68a8d into main May 9, 2026
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.

2 participants