replace GPU 1./sqrt with rsqrt#1741
Merged
wanghan-iapcm merged 3 commits intodeepmodeling:develfrom Jun 7, 2022
Merged
Conversation
Per NVIDIA doc: > 11.1.3. Reciprocal Square Root > The reciprocal square root should always be invoked explicitly as rsqrtf() for single precision and rsqrt() for double precision. The compiler optimizes 1.0f/sqrtf(x) into rsqrtf() only when this does not violate IEEE-754 semantics. See https://docs.nvidia.com/cuda/cuda-c-best-practices-guide/index.html#reciprocal-square-root
Codecov Report
@@ Coverage Diff @@
## devel #1741 +/- ##
=======================================
Coverage 76.14% 76.14%
=======================================
Files 96 96
Lines 7939 7939
=======================================
Hits 6045 6045
Misses 1894 1894 Continue to review full report at Codecov.
|
Member
Author
|
The unit test with CUDA passed as expected. |
denghuilu
approved these changes
Jun 6, 2022
Member
denghuilu
left a comment
denghuilu
left a comment
There was a problem hiding this comment.
The unit test with ROCm also passed as expected.
mingzhong15
pushed a commit
to mingzhong15/deepmd-kit
that referenced
this pull request
Jan 15, 2023
* replace 1./sqrt with rsqrt Per NVIDIA doc: > 11.1.3. Reciprocal Square Root > The reciprocal square root should always be invoked explicitly as rsqrtf() for single precision and rsqrt() for double precision. The compiler optimizes 1.0f/sqrtf(x) into rsqrtf() only when this does not violate IEEE-754 semantics. See https://docs.nvidia.com/cuda/cuda-c-best-practices-guide/index.html#reciprocal-square-root * remove FPTYPE as it has been FPTYPE * apply the same opt for ROCM
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
Per NVIDIA doc:
See https://docs.nvidia.com/cuda/cuda-c-best-practices-guide/index.html#reciprocal-square-root