[
https://issues.apache.org/jira/browse/ARROW-8936?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17366902#comment-17366902
]
Wes McKinney commented on ARROW-8936:
-------------------------------------
Yes, I think it's still useful. Multiple query engine operators will need to
evaluate expressions, so if they want to e.g. break a large chunk (32K -> 1M
rows or more) into smaller pieces (1K -> 32K rows) to parallelize execution, it
would be useful to centralize this implementation. This would need to compose
with the research work happening on work stealing that [~westonpace] has been
working on
> [C++] Parallelize execution of arrow::compute::ScalarFunction
> -------------------------------------------------------------
>
> Key: ARROW-8936
> URL: https://issues.apache.org/jira/browse/ARROW-8936
> Project: Apache Arrow
> Issue Type: New Feature
> Components: C++
> Reporter: Wes McKinney
> Assignee: Wes McKinney
> Priority: Major
> Fix For: 5.0.0
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
