Skip to content

Conversation

@joseph-isaacs
Copy link
Contributor

@joseph-isaacs joseph-isaacs commented Feb 2, 2026

Enough to run a tpch-h projection on the GPU and move it back to the CPU and return to datafusion to show e2e. Likely very slow.

Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
Signed-off-by: Joe Isaacs <joe.isaacs@live.co.uk>
@joseph-isaacs joseph-isaacs added the do not merge Pull requests that are not intended to merge label Feb 2, 2026
@joseph-isaacs joseph-isaacs changed the title wip wip[cuda]: datafusion backed GPU scan Feb 2, 2026
@codspeed-hq
Copy link

codspeed-hq bot commented Feb 2, 2026

CodSpeed Performance Report

Merging this PR will degrade performance by 33.03%

Comparing ji/df-cuda-scan (fdd9e5e) with develop (941edaa)

Summary

❌ 1 regressed benchmark
✅ 1137 untouched benchmarks
⏩ 1384 skipped benchmarks1

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

Mode Benchmark BASE HEAD Efficiency
Simulation chunked_bool_into_canonical[(1000, 10)] 43 µs 64.2 µs -33.03%

Footnotes

  1. 1384 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports.

async move {
let chunk = chunk_result?;

// Execute on CUDA - this is the main difference from CPU path
Copy link

Choose a reason for hiding this comment

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

This is quite a beautiful expression of decoding/filtering on the GPU

Copy link
Contributor Author

Choose a reason for hiding this comment

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

haha

Copy link

Choose a reason for hiding this comment

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

But seriously, as @lwwmanning says, this is a sign that you have your APIs designed nicely

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

do not merge Pull requests that are not intended to merge

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants