oxc/benchmark/README.md

Benchmarks

Initialize

./init.sh

Bench

./bench.sh

Results

The benchmark uses the webpack/lib directory, which contains 537 lintable files.

CPU

• machdep.cpu.brand_string: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz
• machdep.cpu.core_count: 6

Single Run

Using the shell time command for cpu utilization

• oxc: 638% cpu
• ESLint: 161% cpu

Hyperfine

Intel i7 6-cores

Benchmark 1: oxc
  Time (mean ± σ):      36.0 ms ±   3.2 ms    [User: 166.0 ms, System: 69.2 ms]
  Range (min … max):    32.4 ms …  51.8 ms    63 runs

Benchmark 2: eslint
  Time (mean ± σ):      3.044 s ±  0.187 s    [User: 4.606 s, System: 0.260 s]
  Range (min … max):    2.824 s …  3.387 s    10 runs

Summary
  'oxc' ran
   84.66 ± 9.18 times faster than 'eslint'

M2 8-cores

Benchmark 1: oxc
  Time (mean ± σ):      22.0 ms ±   1.0 ms    [User: 101.3 ms, System: 23.2 ms]
  Range (min … max):    20.3 ms …  26.8 ms    122 runs

Benchmark 2: eslint
  Time (mean ± σ):      1.413 s ±  0.014 s    [User: 2.435 s, System: 0.105 s]
  Range (min … max):    1.396 s …  1.438 s    10 runs

Summary
  'oxc' ran
   64.27 ± 2.93 times faster than 'eslint'

Update:

• Rome has been removed from the benchmark because the rules are not the same anymore