YJIT Benchmarks

Details for Benchmarks at 2021-10-08 19:12:29

Overall YJIT is 27.7% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 24.1% 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 MJIT3.0 warmups MJIT3.0 iters YJIT warmups YJIT iters
activerecord 10 130 10 123 10 177
hexapdf 10 10 10 10 10 10
liquid-render 10 107 10 114 10 142
mail 10 116 10 123 10 134
psych-load 10 10 10 10 10 10
railsbench 10 10 10 10 10 10
binarytrees 10 45 10 69 10 58
fannkuchredux 10 10 10 10 10 10
jekyll 10 10 10 10 10 10
lee 10 10 10 13 10 16
nbody 10 172 10 330 10 257
optcarrot 10 10 10 10 10 10
30k_ifelse 10 10 10 10 10 53
30k_methods 10 10 10 10 10 17
cfunc_itself 10 204 10 334 10 508
fib 10 91 10 303 10 375
getivar 10 212 10 211 10 529
keyword_args 10 77 10 98 10 75
respond_to 10 73 10 90 10 108
setivar 10 282 10 264 10 514

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) 1109850 863386 185 1146 0 0% 0 0 0
hexapdf (click) 973953 812761 603 7853 675 8% 0 0 320
liquid-render (click) 455508 371371 143 1245 94 7% 0 0 1
mail (click) 991002 766033 337 5210 129 2% 0 0 39
psych-load (click) 184362 148559 56 408 1 0% 0 0 0
railsbench (click) 2566961 1997701 1293 9591 262 2% 0 0 26
binarytrees (click) 112893 89319 9 49 0 0% 0 0 0
fannkuchredux (click) 129533 103338 6 203 0 0% 0 0 0
jekyll (click) 1954010 1624518 336 3744 173 4% 0 0 0
lee (click) 234602 180481 85 1086 114 10% 1 0 13
nbody (click) 123197 97821 8 149 0 0% 0 0 0
optcarrot (click) 649981 584582 286 5575 20 0% 0 0 0
30k_ifelse (click) 5561021 4890642 9261 57795 0 0% 0 0 0
30k_methods (click) 2117949 1816837 5780 19352 0 0% 0 0 0
cfunc_itself (click) 112253 88530 6 40 0 0% 0 0 0
fib (click) 110717 88103 6 29 0 0% 0 0 0
getivar (click) 112125 90566 6 56 0 0% 0 0 0
keyword_args (click) 110141 87283 7 22 0 0% 0 0 0
respond_to (click) 114173 89696 6 55 0 0% 0 0 0
setivar (click) 112445 89253 6 29 0 0% 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.