Skip to content

improve type stability and runtime performance - #194

Merged
piever merged 2 commits into
JuliaArrays:masterfrom
ranocha:hr/performance
Aug 25, 2021
Merged

piever merged 2 commits into
JuliaArrays:masterfrom
ranocha:hr/performance

Conversation

@ranocha

@ranocha ranocha commented Aug 24, 2021

Copy link
Copy Markdown
Contributor

This is part of an effort to improve the performance of some of our numerical simulation methods for partial differential equations in Trixi.jl. The changes in this PR improve the performance of our methods by ca. 25%. Thus, it would be really great if this could be merged and a new version of StructArrays.jl could be registered.

@piever

piever commented Aug 25, 2021

Copy link
Copy Markdown
Collaborator

Thanks for the PR! Did you happen to measure if the compilation overhead when doing using StructArrays after this change compared with master?

@ranocha

ranocha commented Aug 25, 2021

Copy link
Copy Markdown
Contributor Author

Running julia -e '@time using StructArrays' with Julia 1.7.0-beta3.0 (2021-07-07) several times yields the following results.

  • Current master: A typical result is 0.533961 seconds (2.59 M allocations: 182.671 MiB, 3.20% gc time, 0.88% compilation time); the times vary slightly by ca. 0.02 seconds
  • This PR: A typical result is 0.536773 seconds (2.59 M allocations: 182.672 MiB, 3.23% gc time, 0.88% compilation time); the times vary slightly by ca. 0.02 seconds

Thus, the latency of using StructArrays did not really change.

@piever
piever merged commit 9db8de0 into JuliaArrays:master Aug 25, 2021
@piever piever mentioned this pull request Sep 4, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants