YJIT Benchmarks

Details for Benchmarks at 2022-03-26 19:12:33

Overall YJIT is 32.4% faster than interpreted CRuby!
On Railsbench specifically, YJIT is 31.6% 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 5 155 75 155 20 155
mail 5 124 75 124 20 124
psych-load 5 15 75 15 20 15
railsbench 5 15 75 15 20 15
binarytrees 5 66 75 66 20 66
chunky_png 5 30 75 30 20 30
discourse 5 37 75 37 20 37
erubi 5 47 75 47 20 47
erubi_rails 5 987 75 987 20 987
fannkuchredux 5 15 74 15 20 15
lee 5 20 75 20 20 20
nbody 5 283 75 283 20 283
optcarrot 5 15 75 15 20 15
rubykon 5 15 23 15 20 15
30k_ifelse 5 56 75 56 20 56
30k_methods 5 22 29 22 20 22
cfunc_itself 5 444 75 444 20 444
fib 5 320 75 320 20 320
getivar 5 806 75 806 20 806
keyword_args 5 376 75 376 20 376
respond_to 5 101 75 101 20 101
setivar 5 1824 75 1824 20 1824

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) 985041 753924 185 1199 0 0% 0 0 0
hexapdf (click) 1114897 913437 599 8165 615 7% 0 0 320
liquid-render (click) 489041 388411 143 1507 95 6% 0 0 1
mail (click) 846545 636558 324 5011 125 2% 0 0 39
psych-load (click) 331153 261768 59 434 1 0% 0 0 0
railsbench (click) 2532433 1929142 1297 9350 268 2% 0 0 26
binarytrees (click) 152785 117243 11 59 0 0% 0 0 0
chunky_png (click) 366481 285853 85 864 0 0% 0 0 0
discourse (click) 23540049 11711192 3093 299150 276059 92% 2 0 133
erubi (click) 322641 255427 10 79 0 0% 0 0 0
erubi_rails (click) 1906001 1434484 262 1890 18 0% 0 0 3
fannkuchredux (click) 158737 120555 8 189 0 0% 0 0 0
lee (click) 372945 293715 45 587 69 11% 0 0 12
nbody (click) 162193 124557 10 159 0 0% 0 0 0
optcarrot (click) 489809 430403 198 3576 20 0% 0 0 0
rubykon (click) 293841 237410 140 1533 1 0% 0 0 0
30k_ifelse (click) 5541841 4337567 9263 57805 0 0% 0 0 0
30k_methods (click) 2157009 1646029 5782 19362 0 0% 0 0 0
cfunc_itself (click) 143441 108486 8 50 0 0% 0 0 0
fib (click) 149841 115480 8 39 0 0% 0 0 0
getivar (click) 151569 117991 8 66 0 0% 0 0 0
keyword_args (click) 152145 116513 9 52 0 0% 0 0 0
respond_to (click) 153425 117326 8 65 0 0% 0 0 0
setivar (click) 152081 116844 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.