Skip to content

Conversation

brandtbucher
Copy link
Member

@brandtbucher brandtbucher commented Nov 22, 2024

This should be covered by the NEWS entry in GH-126816. Looks like a ~0.5% speedup and ~0.5% memory savings, so nothing too dramatic there. Only 1/3 as many traces created though, with no change to the number of uops executed, so the memory and time we're saving does seem like it's completely wasted currently (stats).

I'm running benchmarks across all platforms now.

@brandtbucher brandtbucher added performance Performance or resource usage skip news interpreter-core (Objects, Python, Grammar, and Parser dirs) 3.14 bugs and security fixes topic-JIT labels Nov 22, 2024
@brandtbucher brandtbucher self-assigned this Nov 22, 2024
@brandtbucher
Copy link
Member Author

Results across platforms. Very similar to GH-126816, in that it looks like the memory savings are more pronounced on AArch64 macOS and the performance impact is more pronounced on AArch64 Linux.

  • aarch64-apple-darwin: 0.6% faster, 1.1% less memory
  • aarch64-unknown-linux-gnu: 7.0% faster, 0.0% less memory
  • x86_64-unknown-linux-gnu: 0.3-0.6% faster, 0.4-0.5% less memory
  • x86_64-pc-windows-msvc: 0.3% slower
  • i686-pc-windows-msvc: 1.1% faster

Another nice win for Arm, @diegorusso. :)

@brandtbucher brandtbucher merged commit 94b8f8b into python:main Dec 4, 2024
48 of 49 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

3.14 bugs and security fixes interpreter-core (Objects, Python, Grammar, and Parser dirs) performance Performance or resource usage skip news topic-JIT

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant