YJIT Benchmarks

Details for Benchmarks at 2023-01-13 20:10:13

Overall YJIT is 36.5% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 39.0% 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 5 177 5 177
hexapdf 5 15 5 15
liquid-render 5 180 5 180
mail 5 138 5 138
psych-load 5 15 5 15
railsbench 5 15 5 15
ruby-lsp 5 286 5 286
sequel 5 134 5 134
binarytrees 5 87 5 87
chunky_png 5 32 5 32
erubi 5 67 5 67
erubi_rails 5 1164 5 1164
etanni 5 42 5 42
fannkuchredux 5 22 5 22
lee 5 21 5 21
nbody 5 280 5 280
optcarrot 5 15 5 15
rubykon 5 15 5 15
30k_ifelse 5 50 5 50
30k_methods 5 21 5 21
cfunc_itself 5 486 5 486
fib 5 325 5 325
getivar 5 437 5 437
keyword_args 5 384 5 384
respond_to 5 702 5 702
setivar 5 1360 5 1360
setivar_object 5 421 5 421
str_concat 5 455 5 455

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) 2280665 2278916 22 377 8 2% 0 0 0
hexapdf (click) 1721535 1720716 820 10441 217 2% 0 0 0
liquid-render (click) 569303 567690 156 1844 32 1% 0 0 0
mail (click) 1248016 1246616 373 4663 176 3% 0 0 0
psych-load (click) 782640 780958 70 651 19 2% 0 0 0
railsbench (click) 5535004 5534540 1779 14345 483 3% 0 0 0
ruby-lsp (click) 5400540 5398618 6428 41809 1538 3% 14404 0 0
sequel (click) 1454876 1452966 17 428 5 1% 0 0 0
binarytrees (click) 96244 93982 16 140 3 2% 0 0 0
chunky_png (click) 803872 803296 89 1192 31 2% 0 0 0
erubi (click) 741488 739920 13 149 4 2% 0 0 0
erubi_rails (click) 4666847 4666560 329 2830 111 3% 0 0 0
etanni (click) 144397 142490 13 108 3 2% 0 0 0
fannkuchredux (click) 107885 107242 14 273 3 1% 0 0 0
lee (click) 805947 804660 79 820 24 2% 0 0 0
nbody (click) 109125 108094 15 245 5 2% 0 0 0
optcarrot (click) 520808 518938 203 4449 92 2% 0 0 0
rubykon (click) 235505 233144 145 1640 30 1% 0 0 0
30k_ifelse (click) 5245189 5242908 9268 57200 3 0% 0 0 0
30k_methods (click) 1993463 1991910 5787 19427 3 0% 0 0 0
cfunc_itself (click) 93744 92254 13 123 3 2% 0 0 0
fib (click) 92325 91458 13 112 3 2% 0 0 0
getivar (click) 94088 92590 13 139 3 2% 0 0 0
keyword_args (click) 94528 92982 14 125 3 2% 0 0 0
respond_to (click) 94250 92738 13 139 4 2% 0 0 0
setivar (click) 93125 91870 13 119 3 2% 0 0 0
setivar_object (click) 93457 92018 13 119 3 2% 0 0 0
str_concat (click) 93983 92570 15 138 5 3% 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.