YJIT Benchmarks

Details for Benchmarks at 2022-11-21 20:10:19

Overall YJIT is 33.7% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 32.4% faster than CRuby!

The basic "faster" measurement is the geomean of all "headlining" x86 benchmarks on this page.

Headlining Benchmarks

These are "headlining" because the "overall" speedup above is based on these benchmarks specifically.

Speed of each Ruby implementation (iterations/second) relative to the CRuby interpreter. Higher is better.

Other Benchmarks

Speed of each Ruby implementation (iterations/second) relative to the CRuby interpreter. Higher is better.

MicroBenchmarks

Speed of each Ruby implementation (iterations/second) relative to the CRuby interpreter. Higher is better.

Want Raw Graphs and CSV?

Benchmarks Speed Details

Benchmark Memory Usage Details

Number of Iterations and Warmups Tested

bench No JIT warmups No JIT iters YJIT warmups YJIT iters
activerecord 5 209 5 209
hexapdf 5 15 5 15
liquid-render 5 172 5 172
mail 5 123 5 123
psych-load 5 15 5 15
railsbench 5 15 5 15
ruby-lsp 5 277 5 277
binarytrees 5 84 5 84
chunky_png 5 32 5 32
erubi 5 69 5 69
erubi_rails 5 1121 5 1121
etanni 5 43 5 43
fannkuchredux 5 15 5 15
lee 5 21 5 21
nbody 5 225 5 225
optcarrot 5 15 5 15
rubykon 5 15 5 15
30k_ifelse 5 50 5 50
30k_methods 5 20 5 20
cfunc_itself 5 496 5 496
fib 5 327 5 327
getivar 5 427 5 427
keyword_args 5 385 5 385
respond_to 5 634 5 634
setivar 5 410 5 410
str_concat 5 479 5 479

Different Ruby configurations want different amounts of warmup. With no JIT, CRuby needs hardly any. YJIT and MJIT 3.0 both warm up quite quickly, while MJIT in 3.1 often slows down for a time as it compiles, after an unpredictable delay.

Benchmark YJIT Stats

Hover your cursor over the benchmark names for descriptions of each benchmark.

bench Exit Report Inline Outlined Comp iSeqs Comp Blocks Inval Inval Ratio Bind Alloc Bind Set Const Bumps
activerecord (click) 2290360 2288498 25 383 6 1% 0 0 0
hexapdf (click) 1759122 1757275 819 10151 218 2% 0 0 0
liquid-render (click) 588654 586441 156 1837 34 1% 0 0 0
mail (click) 1326441 1326318 373 5107 176 3% 0 0 0
psych-load (click) 801466 799513 70 633 19 3% 0 0 0
railsbench (click) 5609893 5609332 1779 13938 483 3% 0 0 0
ruby-lsp (click) 5427979 5427647 6425 40537 1520 3% 14004 0 0
binarytrees (click) 136253 134679 16 139 4 2% 0 0 0
chunky_png (click) 820750 820265 89 1173 31 2% 0 0 0
erubi (click) 746241 745943 13 147 4 2% 0 0 0
erubi_rails (click) 4729807 4728266 329 2686 111 4% 0 0 0
etanni (click) 135436 134074 13 107 4 3% 0 0 0
fannkuchredux (click) 146721 144465 13 269 4 1% 0 0 0
lee (click) 824819 822991 79 804 25 3% 0 0 0
nbody (click) 148414 147793 15 248 6 2% 0 0 0
optcarrot (click) 557446 556977 203 4133 93 2% 0 0 0
rubykon (click) 281068 280097 146 1656 32 1% 0 0 0
30k_ifelse (click) 5404118 5403616 9268 57196 4 0% 0 0 0
30k_methods (click) 2094589 2092547 5787 19425 4 0% 0 0 0
cfunc_itself (click) 132582 132096 13 122 4 3% 0 0 0
fib (click) 131127 131044 13 111 4 3% 0 0 0
getivar (click) 133373 132581 13 138 4 2% 0 0 0
keyword_args (click) 133371 132581 14 124 4 3% 0 0 0
respond_to (click) 133502 132581 13 138 5 3% 0 0 0
setivar (click) 133365 132513 13 111 4 3% 0 0 0
str_concat (click) 134383 133173 15 137 6 4% 0 0 0

YJIT stats correspond to the YJIT stats exit report.

Note: currently, all stats are collected on x86_64, not ARM.

Raw JSON data files

All graphs and table data in this page comes from processing these data files, which come from benchmark runs.


Far Too Many Details

These are links to HTML reports on GitHub with far, far too much detail about the benchmarking runs per-platform.


Autogenerated by continuous_reporting/generate_and_upload_reports.rb script.