Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Remove batch size bytes limits #11746

Merged
merged 4 commits into from
Nov 22, 2024
Merged

Remove batch size bytes limits #11746

merged 4 commits into from
Nov 22, 2024

Conversation

zpuller
Copy link
Collaborator

@zpuller zpuller commented Nov 21, 2024

Testing with higher batch sizes has shown potential performance improvements in certain scenarios, and has not shown any evidence of query failures or data corruption, so we think we are comfortable removing the existing limits we've been imposing on these configs.

Signed-off-by: Zach Puller <[email protected]>
Signed-off-by: Zach Puller <[email protected]>
@@ -223,8 +223,6 @@ case class RequireSingleBatchWithFilter(filterExpression: GpuExpression)
case class TargetSize(override val targetSizeBytes: Long)
Copy link
Collaborator Author

Choose a reason for hiding this comment

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

not sure if we still need targetSizeBytes

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

I think CoalesceSizeGoal still needs it

Copy link
Member

Choose a reason for hiding this comment

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

Yes, we still need targetSizeBytes.

Signed-off-by: Zach Puller <[email protected]>
Signed-off-by: Zach Puller <[email protected]>
@sameerz sameerz added the performance A performance related task/issue label Nov 22, 2024
@abellina
Copy link
Collaborator

build

Copy link
Collaborator

@revans2 revans2 left a comment

Choose a reason for hiding this comment

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

We probably should look through our documentation to be sure that this limit is not mentioned anywhere. But some docs are not in this repo.

@zpuller zpuller merged commit e5547a1 into NVIDIA:branch-24.12 Nov 22, 2024
49 checks passed
@zpuller zpuller deleted the batch branch December 31, 2024 17:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
performance A performance related task/issue
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants