YJIT Benchmarks

Details for Benchmarks at 2021-10-27 07:09:55

Overall YJIT is 28.3% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 26.5% 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 10 131 10 121 10 175
hexapdf 10 10 10 10 10 10
liquid-render 10 107 10 113 10 149
mail 10 117 10 125 10 135
psych-load 10 10 10 10 10 10
railsbench 10 10 10 10 10 10
binarytrees 10 44 10 67 10 57
fannkuchredux 10 10 10 10 10 10
jekyll 10 10 10 10 10 10
lee 10 10 10 13 10 16
nbody 10 176 10 331 10 256
optcarrot 10 10 10 10 10 10
rubykon 10 10 10 10 10 10
30k_ifelse 10 10 10 10 10 53
30k_methods 10 10 10 10 10 17
cfunc_itself 10 205 10 339 10 489
fib 10 92 10 308 10 375
getivar 10 214 10 211 10 529
keyword_args 10 73 10 97 10 429
respond_to 10 76 10 90 10 108
setivar 10 275 10 262 10 451

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) 1116122 870729 185 1145 0 0% 0 0 0
hexapdf (click) 1073012 897501 603 7827 619 7% 0 0 320
liquid-render (click) 458177 374973 143 1247 95 7% 0 0 1
mail (click) 833716 645681 339 5013 124 2% 0 0 39
psych-load (click) 193194 156510 56 409 1 0% 0 0 0
railsbench (click) 2573617 2010493 1293 9479 262 2% 0 0 26
binarytrees (click) 114410 90765 9 48 0 0% 0 0 0
fannkuchredux (click) 130474 104390 6 202 0 0% 0 0 0
jekyll (click) 1966618 1635834 336 3785 173 4% 0 0 0
lee (click) 235434 181947 85 1082 114 10% 1 0 13
nbody (click) 124266 98741 8 148 0 0% 0 0 0
optcarrot (click) 460394 430533 196 3566 20 0% 0 0 0
rubykon (click) 255914 221191 138 1506 1 0% 0 0 0
30k_ifelse (click) 5562090 4891562 9261 57794 0 0% 0 0 0
30k_methods (click) 2118570 1817511 5780 19351 0 0% 0 0 0
cfunc_itself (click) 113258 89450 6 39 0 0% 0 0 0
fib (click) 111402 88645 6 28 0 0% 0 0 0
getivar (click) 113194 91486 6 55 0 0% 0 0 0
keyword_args (click) 113706 89960 7 41 0 0% 0 0 0
respond_to (click) 115114 90748 6 54 0 0% 0 0 0
setivar (click) 113130 89927 6 28 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.