-
Notifications
You must be signed in to change notification settings - Fork 137
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
[LIBM] Introduce faster method for evaluating polynomials #239
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
Nice stuff. Could you please provide a speedup report on both Intel and AArch64? |
fpetrogalli
requested changes
Feb 11, 2019
fpetrogalli
approved these changes
Feb 12, 2019
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Outdated suggestions cannot be applied.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.
This patch replaces Horner method which was used to evaluate polynomials with Estrin's method( https://en.wikipedia.org/wiki/Estrin%27s_scheme ) that allows more parallel computations with out-of-order execution.
This patch also introducing a new reduction method to tan.
With this patch, mainly computation for double-precision functions becomes faster, and the effect is like a few percent to 20 percent. For example, the ratio between execution time of the following functions before and after applying this patch is shown below.
Sleef_atan2d4_u35 : 1.21
Sleef_powd4_u10 : 1.17
Sleef_sind4_u35 : 1.10
Sleef_tand4_u10 : 1.04
Sleef_tand4_u35 : 1.17