This repository has been archived by the owner on Mar 21, 2024. It is now read-only.
Remove policies for arch < sm35. #213
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
|
DVS CL 29199322 |
alliepiper
added
the
testing: internal ci in progress
griwes
approved these changes
Oct 14, 2020
alliepiper
removed
the
testing: internal ci in progress
alliepiper
force-pushed
the
enh/remove_old_sm_archs
branch
from
8b3b747 to
fc38233
Compare
|
Starting DVS on CL 29260540. 🤞 |
alliepiper
added
the
testing: internal ci in progress
|
No timeouts 🎉 |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.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.
Goal is to clean up code and reduce build time:
Before:
After:
Thrust isn't significantly affected, as it spends most of its time building non-CUB algorithms (the set operation and minmax tests in particular). We can probably bring that down by removing old SM policies in a follow up patch, though Thrust currently only appears to go down to sm30 so there are fewer policies to take out.
CUB's compile time is reduced by nearly 40%. There's still a lot of room for improvement by looking at the CPU vs walltime (build used 12 cores). This is because there are one or two tests that take a very long time to compile. We can improve this by splitting up those tests so they can be parallelized.