CI: CodSpeed w/ AVX2 - #1634
Merged
Merged
Conversation
AVX2 is a subset of the CPU features we know exist in `-march=x86-64-v3` and are always present on GH action x86-64 cloud runners. Confuses measurements to compile for them and disable at runtime.
ax3l
added
Performance optimization
component: tests
ax3l
commented
Aug 25, 2026
Some of the newer GH runners support AVX152 Co-authored-by: Axel Huebl <axel.huebl@plasma.ninja>
Merging this PR will improve performance by 11.05%
|
| Benchmark | BASE |
HEAD |
Efficiency | |
|---|---|---|---|---|
| ⚡ | test_ExactCFbend[spin] |
48.5 ms | 41.2 ms | +17.65% |
| ⚡ | test_PolygonAperture[nospin] |
8.9 ms | 7.7 ms | +16.39% |
| ⚡ | test_PolygonAperture[spin] |
8.9 ms | 7.7 ms | +16.38% |
| ⚡ | test_ExactCFbend[nospin] |
23.2 ms | 19.9 ms | +16.24% |
| ⚡ | test_NonlinearLens[nospin] |
1.4 ms | 1.3 ms | +12.61% |
| ⚡ | test_NonlinearLens[spin] |
2.6 ms | 2.4 ms | +10.35% |
| ⚡ | test_Sol[spin] |
2.2 ms | 2 ms | +9.34% |
| ⚡ | test_Sbend[spin] |
2.1 ms | 1.9 ms | +9.18% |
| ⚡ | test_ExactQuad[nospin] |
5.2 ms | 4.9 ms | +7.83% |
| ⚡ | test_space_charge_solver[3D_fft] |
330.1 ms | 307.6 ms | +7.32% |
| ⚡ | test_space_charge_solver[2p5D_fft] |
195.7 ms | 182.6 ms | +7.13% |
| ⚡ | test_SoftQuadrupole[spin] |
823.5 µs | 769.1 µs | +7.07% |
| ⚡ | test_space_charge_solver[2D_fft] |
197.4 ms | 184.4 ms | +7.06% |
Tip
Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.
Comparing ax3l:ci-avx2-codspeed (3f05481) with development (92ccb0d)
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
AVX2 is a subset of the CPU features we know exist in
-march=x86-64-v3and are always present on GH action x86-64 cloud runners. Confuses measurements to compile for them and disable at runtime.Seen as an issue in #1632
AVX2 has been on every Intel Xeon since Haswell (2013) and every AMD EPYC since Zen 1 (2017), which covers every CPU family Azure has ever backed these runners with.