Update strings contains benchmarks to nvbench #15495
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
davidwendt
added
2 - In Progress
mythrocks
reviewed
Apr 16, 2024
| d_target.empty()); | ||
|
|
||
| if (!d_target.empty()) { | ||
| if (d_target.empty()) { |
There was a problem hiding this comment.
👍 This didn't occur to me, and I've been staring at this code for a bit now.
3 tasks
|
This also includes some performance improvement for long strings From #15536 (comment) |
davidwendt
added
3 - Ready for Review
|
Adding a link here to the code logic attempts for reference and convenience |
PointKernel
requested changes
May 13, 2024
ttnghia
reviewed
May 16, 2024
cpp/benchmarks/string/find.cpp
Outdated
Comment on lines
66
to
77
| state.exec(nvbench::exec_tag::sync, [&](nvbench::launch& launch) { | ||
| if (api == "find") { | ||
| cudf::strings::find(input, target); | ||
| } else if (api == "find_multi") { | ||
| cudf::strings::find_multiple(input, cudf::strings_column_view(targets)); | ||
| } else if (api == "contains") { | ||
| cudf::strings::contains(input, target); | ||
| } else if (api == "starts_with") { | ||
| cudf::strings::starts_with(input, target); | ||
| } else if (api == "ends_with") { | ||
| cudf::strings::ends_with(input, target); | ||
| } |
There was a problem hiding this comment.
Will multiple if/else inside one state.exec lambda affect the benchmark run (as this will be executed in thousands of iterations)?
How about placing separate state.exec in each if/else branch?
ttnghia
reviewed
May 16, 2024
Comment on lines
+373
to
+376
| if (((i + j + d_target.size_bytes()) <= d_str.size_bytes()) && | ||
| d_target.compare(d_str.data() + i + j, d_target.size_bytes()) == 0) { | ||
| found = true; | ||
| } |
There was a problem hiding this comment.
Will any early termination be helpful here?
There was a problem hiding this comment.
I benchmarked that and it did not help -- technically a bit slower.
ttnghia
approved these changes
May 16, 2024
|
/merge |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
Reference #15405
Updates the benchmarks for
cudf::strings::contains()to use nvbench and also introduce a hit-test axis.The logic has been updated to remove the unneeded
fill()call for long strings.Also cleaned up code and updated logic to process 4 bytes per warp thread.
Checklist