Skip to content

Conversation

sandrathena
Copy link

implemented FIXME in spark instruction

@sandrathena
Copy link
Author

Here are our experiment diagrams: rowcumsumcumsum1500
rowcumsumcumsum3500
rowcumsumtrans1500
rowcumsumtrans3500

@sandrathena
Copy link
Author

Here are our detailed experiment diagrams: cumsum_1500
cumsum_3500
transposedcumsum_1500
transposedcumsum_3500

@mboehm7
Copy link
Contributor

mboehm7 commented Aug 3, 2025

Thanks for sharing theses experimental results @sandrathena. The results look good and are absolutely in the range what we would expect - the core rowcumsum is slightly faster because of sequential access to the row-major layout, and the rewrite avoids the costly transpose.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
Status: In Progress
Development

Successfully merging this pull request may close these issues.

2 participants