Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
evnwashere
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
1.
▲
by
evnwashere
2y ago
That’s why i created mitata, it greatly improves on javascript (micro-)benchmarking tooling it provides bunch of features to help avoiding jit optimization foot-guns during benchmarking and dips into more advanced stuff like hardware cpu co
2.
▲
by
evnwashere
2y ago
vitest is nice but it’s completely unsuited for micro-benchmarks as it ends up oom crashing after just 2 optimized out benchmarks
3.
▲
by
evnwashere
2y ago
i didn’t want it to be complex so it uses simple time budget + at least x amount of samples, both and more can be configured with lower level api. in practice i haven’t found any js function that gets faster after mitata’s time budget (excl
4.
▲
by
evnwashere
2y ago
it was hand picked for exactly that :)
5.
▲
by
evnwashere
2y ago
yes, you can import mitata inside web worker and run it there, if you only need raw results you can even use lower level api https://github.com/evanwashere/mitata?tab=readme-ov-file#giv...
6.
▲
by
evnwashere
2y ago
I have been thinking of reusing/creating something like https://perf.rust-lang.org/ that lets you pick and compare specific hash/commit with all data from json format
7.
▲
by
evnwashere
2y ago
It works anywhere where javascript works, so you can easily run it in browser too. Tho idea of making jsbench like website but with mitata accuracy (+ dedicated runners) keeps bugging me.
8.
▲
Show HN: Mitata – Benchmarking tooling for JavaScript
(github.com)
93 points
by
evnwashere
2y ago
|
25 comments