Refresh engine comparison benchmark results#2598
Merged
Merged
Conversation
Re-run after the builtin-shapes rollout and the 2026-07 performance campaign (sebastienros#2580-sebastienros#2597). Highlights vs the 4.11.0 refresh: minimal script -47% time / -45% allocation (prepared execution now under a microsecond), evaluation -33% allocation, stopwatch-modern -22% (ahead of NiL.JS, 1.13x behind Jurassic, from 1.44x), object-string -14%, string-base64 -12% (taking the lead from NiL.JS), object-array -11%. Competitor versions unchanged; their rows served as thermal canaries. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01S2U2y9voMQV9rvCkYeQbka
|
Amazing work, Marko! I've been following the work on Jint for about 10 years and it's never been more exciting to see the progress than right now. I didn't know the right place to express my gratitude, so forgive the noise, but thank you for all your hard work! |
This was referenced Jul 13, 2026
This was referenced Jul 20, 2026
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Re-run of
EngineComparisonBenchmark(full suite, BenchmarkDotNet default jobs, same hardware/SDK as the previous refresh) after the builtin-shapes rollout and the 2026-07 performance campaign (#2580–#2597). Competitor versions unchanged — their rows served as thermal canaries and read within ±3% of the previous refresh.Highlights vs the 4.11.0 refresh (prepared-script path)
minimalevaluation/-modernstopwatch-moderndromaeo-object-stringdromaeo-string-base64/-moderndromaeo-object-arraydromaeo-3d-cubeRank changes:
dromaeo-string-base64— Jint took the lead from NiL.JS (both variants);stopwatch-modern— Jint moved ahead of NiL.JS and closed on Jurassic (1.44× → 1.13×);minimalsource mode now also beats NiL.JS.One row reads slower:
dromaeo-object-array-modern+10%, a family with a documented ±7–11% run-to-run swing whose classic sibling improved 11%. The reading reproduces identically on builds with and without the most recent merges (verified via a pre-#2595 worktree A/B), so it is noted in the README as tracked measurement noise pending the next refresh.dromaeo-object-regexp's apparent +15% did not reproduce on re-run (112 ms vs the previous 110 ms; that row swings ~12% between runs).🤖 Generated with Claude Code
https://claude.ai/code/session_01S2U2y9voMQV9rvCkYeQbka