YJIT Benchmarks

Details for Benchmarks at 2021-10-07 19:09:37

Overall YJIT is 24.4% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 20.4% 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 MJIT3.0 warmups MJIT3.0 iters YJIT warmups YJIT iters
activerecord 10 129 10 120 10 177
hexapdf 10 10 10 10 10 10
liquid-render 10 107 10 113 10 128
mail 10 123 10 125 10 139
psych-load 10 10 10 10 10 10
railsbench 10 10 10 10 10 10
binarytrees 10 42 10 69 10 59
fannkuchredux 10 10 10 10 10 10
jekyll 10 10 10 10 10 10
lee 10 10 10 13 10 15
nbody 10 170 10 332 10 262
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 10 217 10 338 10 498
fib 10 94 10 308 10 375
getivar 10 223 10 211 10 527
keyword_args 10 76 10 97 10 73
respond_to 10 79 10 90 10 113
setivar 10 290 10 262 10 460

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) 1112410 865084 185 1163 0 0% 0 0 0
hexapdf (click) 973569 812555 602 7849 677 8% 0 0 320
liquid-render (click) 451092 368148 143 1203 94 7% 0 0 1
mail (click) 970074 749603 337 5006 129 2% 0 0 39
psych-load (click) 184938 148953 56 408 1 0% 0 0 0
railsbench (click) 2560497 1993088 1293 9545 262 2% 0 0 26
binarytrees (click) 113469 89713 9 49 0 0% 0 0 0
fannkuchredux (click) 129149 103092 6 203 0 0% 0 0 0
jekyll (click) 1956634 1626610 336 3736 173 4% 0 0 0
lee (click) 234602 180481 85 1086 114 10% 1 0 13
nbody (click) 122301 97049 8 149 0 0% 0 0 0
optcarrot (click) 649277 583958 286 5575 20 0% 0 0 0
30k_ifelse (click) 5560701 4890396 9261 57795 0 0% 0 0 0
30k_methods (click) 2117245 1816213 5780 19352 0 0% 0 0 0
cfunc_itself (click) 111549 87906 6 40 0 0% 0 0 0
fib (click) 109821 87331 6 29 0 0% 0 0 0
getivar (click) 112189 90566 6 56 0 0% 0 0 0
keyword_args (click) 109757 86905 7 22 0 0% 0 0 0
respond_to (click) 114493 89889 6 55 0 0% 0 0 0
setivar (click) 111613 88481 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.