YJIT Benchmarks

Details for Benchmarks at 2021-11-26 07:10:45

Overall YJIT is 29.3% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 20.7% 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 MJIT warmups MJIT iters YJIT warmups YJIT iters
activerecord 5 192 20 192 75 192 20 192
hexapdf 5 15 20 15 20 15
liquid-render 5 173 20 173 75 173 20 173
mail 5 131 20 131 75 131 20 131
psych-load 5 15 20 15 75 15 20 15
railsbench 5 15 20 15 75 15 20 15
binarytrees 5 85 20 85 75 85 20 85
erubi 5 49 20 49 75 49 20 49
erubi_rails 5 573 20 573 75 573 20 573
fannkuchredux 5 15 20 15 54 15 20 15
jekyll 5 15 20 15 36 15 20 15
lee 5 24 20 24 75 24 20 24
nbody 5 345 20 345 75 345 20 345
optcarrot 5 15 20 15 75 15 20 15
rubykon 5 15 20 15 43 15 20 15
30k_ifelse 5 64 20 64 75 64 20 64
30k_methods 5 26 20 26 41 26 20 26
cfunc_itself 5 514 20 514 75 514 20 514
fib 5 383 20 383 75 383 20 383
getivar 5 563 20 563 75 563 20 563
keyword_args 5 436 20 436 75 436 20 436
respond_to 5 126 20 126 75 126 20 126
setivar 5 481 20 481 75 481 20 481

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) 956010 719089 185 1164 0 0% 0 0 0
hexapdf (click) 1076330 857292 596 8005 614 7% 0 0 321
liquid-render (click) 468855 364529 143 1450 95 6% 0 0 1
mail (click) 863210 635274 325 5431 124 2% 0 0 39
psych-load (click) 290721 225414 57 423 1 0% 0 0 0
railsbench (click) 2566818 1915355 1293 9691 263 2% 0 0 26
binarytrees (click) 112805 85691 9 48 0 0% 0 0 0
erubi (click) 269409 208842 8 35 0 0% 0 0 0
erubi_rails (click) 2436386 1741913 260 1996 17 0% 0 0 3
fannkuchredux (click) 128293 97352 6 202 0 0% 0 0 0
jekyll (click) 1879082 1494574 336 3933 179 4% 0 0 0
lee (click) 330218 256704 44 562 67 11% 0 0 12
nbody (click) 121701 92313 8 148 0 0% 0 0 0
optcarrot (click) 450661 391852 196 3566 20 0% 0 0 0
rubykon (click) 258149 207832 138 1576 1 0% 0 0 0
30k_ifelse (click) 5501477 4305666 9261 57794 0 0% 0 0 0
30k_methods (click) 2117093 1614477 5780 19351 0 0% 0 0 0
cfunc_itself (click) 110885 83971 6 39 0 0% 0 0 0
fib (click) 109349 83579 6 28 0 0% 0 0 0
getivar (click) 111077 86090 6 55 0 0% 0 0 0
keyword_args (click) 111717 84612 7 41 0 0% 0 0 0
respond_to (click) 112997 85425 6 54 0 0% 0 0 0
setivar (click) 111013 84614 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.