Skip to content

Conversation

@pmattione-nvidia
Copy link
Contributor

@pmattione-nvidia pmattione-nvidia commented Nov 19, 2025

Followup to PR# 19643. This changes it so the parquet decode doesn't fill in null values for anything except string lengths and list offsets. This matches the updated dev guide. Benchmarks show no measurable improvement, but it's still a good idea to do less work, as it will help once other bottlenecks are removed.

This also includes minor changes to simplify thread-out-of-bounds checking in the decode loops.

Checklist

  • I am familiar with the Contributing Guidelines.
  • New or existing tests cover these changes.
  • The documentation is up to date with these changes.

@pmattione-nvidia pmattione-nvidia self-assigned this Nov 19, 2025
@pmattione-nvidia pmattione-nvidia added the Performance Performance related issue label Nov 19, 2025
@pmattione-nvidia pmattione-nvidia requested a review from a team as a code owner November 19, 2025 21:33
@pmattione-nvidia pmattione-nvidia added the improvement Improvement / enhancement to an existing function label Nov 19, 2025
@pmattione-nvidia pmattione-nvidia added the non-breaking Non-breaking change label Nov 19, 2025
@github-actions github-actions bot added the libcudf Affects libcudf (C++/CUDA) code. label Nov 19, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

improvement Improvement / enhancement to an existing function libcudf Affects libcudf (C++/CUDA) code. non-breaking Non-breaking change Performance Performance related issue

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants