YJIT Benchmarks

Details for Benchmarks at 2022-03-14 07:13:08

Overall YJIT is 32.3% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 27.8% 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 MJIT warmups MJIT iters YJIT warmups YJIT iters
activerecord 5 181 75 181 20 181
hexapdf 5 15 20 15
liquid-render 5 157 75 157 20 157
mail 5 122 75 122 20 122
psych-load 5 15 75 15 20 15
railsbench 5 15 75 15 20 15
binarytrees 5 63 75 63 20 63
chunky_png 5 30 75 30 20 30
discourse 5 34 75 34 20 34
erubi 5 49 75 49 20 49
erubi_rails 5 973 75 973 20 973
fannkuchredux 5 15 75 15 20 15
lee 5 22 75 22 20 22
nbody 5 283 75 283 20 283
optcarrot 5 15 75 15 20 15
rubykon 5 15 38 15 20 15
30k_ifelse 5 56 75 56 20 56
30k_methods 5 22 29 22 20 22
cfunc_itself 5 449 75 449 20 449
fib 5 320 75 320 20 320
getivar 5 774 75 774 20 774
keyword_args 5 376 75 376 20 376
respond_to 5 108 75 108 20 108
setivar 5 1897 75 1897 20 1897

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) 988817 757457 185 1199 0 0% 0 0 0
hexapdf (click) 1114513 912980 599 8169 613 7% 0 0 320
liquid-render (click) 489105 388027 143 1505 96 6% 0 0 1
mail (click) 873425 655473 324 5277 125 2% 0 0 39
psych-load (click) 330129 260911 59 434 1 0% 0 0 0
railsbench (click) 2535953 1933371 1297 9358 268 2% 0 0 26
binarytrees (click) 152273 116983 11 59 0 0% 0 0 0
chunky_png (click) 365265 285154 85 858 0 0% 0 0 0
discourse (click) 22699281 11401628 3134 285471 261949 91% 2 0 133
erubi (click) 321617 254838 10 79 0 0% 0 0 0
erubi_rails (click) 1905297 1435392 262 1872 18 0% 0 0 3
fannkuchredux (click) 166545 128050 8 189 0 0% 0 0 0
lee (click) 372753 293979 45 587 69 11% 0 0 12
nbody (click) 161681 124297 10 159 0 0% 0 0 0
optcarrot (click) 488785 429814 198 3576 20 0% 0 0 0
rubykon (click) 293457 237155 140 1532 1 0% 0 0 0
30k_ifelse (click) 5541009 4336733 9263 57805 0 0% 0 0 0
30k_methods (click) 2156881 1645937 5782 19362 0 0% 0 0 0
cfunc_itself (click) 150865 115700 8 50 0 0% 0 0 0
fib (click) 149329 115220 8 39 0 0% 0 0 0
getivar (click) 151057 117731 8 66 0 0% 0 0 0
keyword_args (click) 151633 116253 9 52 0 0% 0 0 0
respond_to (click) 152913 117066 8 65 0 0% 0 0 0
setivar (click) 151057 116255 8 39 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.