YJIT Benchmarks

Details for Benchmarks at 2022-11-14 04:20:06

Overall YJIT is 32.3% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 34.8% 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 5 208 5 208
hexapdf 5 15 5 15
liquid-render 5 175 5 175
mail 5 118 5 118
psych-load 5 15 5 15
railsbench 5 15 5 15
ruby-lsp 5 276 5 276
binarytrees 5 84 5 84
chunky_png 5 31 5 31
erubi 5 67 5 67
erubi_rails 5 1111 5 1111
etanni 5 42 5 42
fannkuchredux 5 15 5 15
lee 5 22 5 22
nbody 5 233 5 233
optcarrot 5 15 5 15
rubykon 5 15 5 15
30k_ifelse 5 49 5 49
30k_methods 5 19 5 19
cfunc_itself 5 483 5 483
fib 5 327 5 327
getivar 5 426 5 426
keyword_args 5 387 5 387
respond_to 5 715 5 715
setivar 5 361 5 361
str_concat 5 453 5 453

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) 2327857 2326888 25 362 6 1% 0 0 0
hexapdf (click) 1798889 1796925 820 10294 189 1% 0 0 0
liquid-render (click) 602535 600827 156 1838 34 1% 0 0 0
mail (click) 1352132 1351572 373 5180 176 3% 0 0 0
psych-load (click) 814419 813236 70 637 19 2% 0 0 0
railsbench (click) 5671091 5669288 1779 13875 484 3% 0 0 0
ruby-lsp (click) 5530204 5528784 6425 40540 1521 3% 13904 0 0
binarytrees (click) 135642 133984 16 137 4 2% 0 0 0
chunky_png (click) 831434 829995 89 1174 31 2% 0 0 0
erubi (click) 752717 750458 13 145 4 2% 0 0 0
erubi_rails (click) 4771526 4769451 329 2684 111 4% 0 0 0
etanni (click) 136490 134687 13 105 4 3% 0 0 0
fannkuchredux (click) 147681 147534 13 267 4 1% 0 0 0
lee (click) 833600 831680 79 803 25 3% 0 0 0
nbody (click) 150291 149244 15 246 6 2% 0 0 0
optcarrot (click) 575956 574916 203 4131 93 2% 0 0 0
rubykon (click) 282414 280740 146 1614 31 1% 0 0 0
30k_ifelse (click) 5427259 5424734 9268 57194 4 0% 0 0 0
30k_methods (click) 2095599 2093105 5787 19423 4 0% 0 0 0
cfunc_itself (click) 133540 132561 13 120 4 3% 0 0 0
fib (click) 132078 131639 13 109 4 3% 0 0 0
getivar (click) 134164 132897 13 136 4 2% 0 0 0
keyword_args (click) 134328 133128 14 122 4 3% 0 0 0
respond_to (click) 134057 132897 13 136 5 3% 0 0 0
setivar (click) 134318 133128 13 109 4 3% 0 0 0
str_concat (click) 126288 125130 15 135 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.