Default to clang++ on aarch64 for faster phase imputation#296
Open
tfenne wants to merge 1 commit into
Open
Conversation
The phase kernels are x86 AVX2 intrinsics that SIMDe lowers to NEON on ARM, and clang generates far better NEON than gcc (~2.5x faster phase kernels on AWS Graviton, bit-identical output). common.mk now defaults CXX to clang++ on aarch64 (falling back to g++ when clang++ is absent) and g++ elsewhere; the Dockerfile installs clang so the aarch64 image picks it up. Installation docs updated accordingly.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
I finally got around to doing some benchmarking of GLIMPSE2_phase on AWS's aarch64/graviton instances. Unlike on mac where we get clang by default, I was generating builds initially with gcc and seeing vastly worse performance. After some digging around and trial and error, it turns out it's all gcc vs. clang. So this PR simply changes the default compiler on all aarch64 platforms to clang, with fallback to gcc.
The phase kernels are x86 AVX2 intrinsics that SIMDe lowers to NEON on ARM, and clang generates far better NEON than gcc (~1.5-2.5x faster phase kernels on AWS Graviton, bit-identical output). common.mk now defaults CXX to clang++ on aarch64 (falling back to g++ when clang++ is absent) and g++ elsewhere; the Dockerfile installs clang so the aarch64 image picks it up. Installation docs updated accordingly.