YJIT Benchmarks

Details for Benchmarks at 2024-02-26 06:08:15 UTC

YJIT metrics from the yjit-bench suite using Ruby 459198999e

Using the geomean of the headline benchmarks for x86 YJIT 3.4.0dev is
  • 71.5% faster than CRuby 3.4.0dev
On railsbench it is
  • 68.2% faster than CRuby 3.4.0dev

Performance on Headline Benchmarks

Speed of each Ruby implementation relative to the baseline CRuby measurement. Higher is better.

Memory Usage on Headline Benchmarks

Memory usage of each Ruby implementation relative to the baseline CRuby measurement. Lower is better.

Performance on Other Benchmarks

Speed of each Ruby implementation relative to the baseline CRuby measurement. Higher is better.

Memory Usage on Other Benchmarks

Memory usage of each Ruby implementation relative to the baseline CRuby measurement. Lower is better.

Performance on MicroBenchmarks

Speed of each Ruby implementation relative to the baseline CRuby measurement. Higher is better.

Memory Usage on MicroBenchmarks

Memory usage of each Ruby implementation relative to the baseline CRuby measurement. Lower is better.

Want Raw Graphs and CSV?

Benchmarks Speed Details

Benchmark Memory Usage Details

Number of Iterations and Warmups Tested

bench CRuby 3.4.0dev warmups CRuby 3.4.0dev iters YJIT 3.4.0dev warmups YJIT 3.4.0dev iters
activerecord 30 512 30 512
chunky-png 30 32 30 32
erubi-rails 30 19 30 19
hexapdf 30 15 30 15
liquid-c 30 346 30 346
liquid-compile 30 341 30 341
liquid-render 30 250 30 250
lobsters 30 23 30 23
mail 30 161 30 161
psych-load 30 15 30 15
railsbench 30 15 30 15
rubocop 30 181 30 181
ruby-lsp 30 184 30 184
sequel 30 289 30 289
binarytrees 30 81 30 81
blurhash 30 101 30 101
erubi 30 77 30 77
etanni 30 51 30 51
fannkuchredux 30 24 30 24
fluentd 30 15 30 15
graphql 30 15 30 15
graphql-native 30 21 30 21
lee 30 20 30 20
matmul 30 18 30 18
nbody 30 315 30 315
nqueens 30 299 30 299
optcarrot 30 15 30 15
rack 30 178 30 178
ruby-json 30 15 30 15
rubykon 30 15 30 15
sudoku 30 26 30 26
tinygql 30 41 30 41
30k_ifelse 30 55 30 55
30k_methods 30 24 30 24
cfunc_itself 30 827 30 827
fib 30 416 30 416
getivar 30 940 30 940
keyword_args 30 591 30 591
respond_to 30 1480 30 1480
ruby-xor 30 96 30 96
setivar 30 2393 30 2393
setivar_object 30 492 30 492
setivar_young 30 507 30 507
str_concat 30 433 30 433
throw 30 747 30 747

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 Compile Time MS
activerecord (click) 890396 771261 49 536 0 0% 0 0 15.730449
chunky-png (click) 337346 291463 87 1096 1 0% 0 0 32.349232
erubi-rails (click) 1725113 1418447 281 2957 6 0% 0 0 84.079456
hexapdf (click) 1426927 1252476 596 12660 38 0% 0 0 368.310544
liquid-c (click) 525996 457583 120 1753 5 0% 0 0 47.027159
liquid-compile (click) 442422 361446 151 2069 2 0% 0 0 62.547048
liquid-render (click) 626276 550937 144 2273 8 0% 0 0 62.7225
lobsters (click) 7897930 6598010 3219 50079 73 0% 0 0 1694.595201
mail (click) 732438 611658 346 4846 14 0% 0 0 141.494428
psych-load (click) 301391 253853 65 655 3 0% 0 0 20.233398
railsbench (click) 3015464 2605190 1643 15681 54 0% 0 0 449.340937
rubocop (click) 5360294 4558793 2902 47734 106 0% 4 0 1419.146791
ruby-lsp (click) 638859 555006 159 2793 14 0% 0 0 77.219767
sequel (click) 498234 430322 16 118 0 0% 0 0 4.332376
binarytrees (click) 9392 7588 11 75 0 0% 0 0 2.967128
blurhash (click) 52000 47236 32 461 0 0% 0 0 15.273546
erubi (click) 271414 222726 10 111 0 0% 0 0 3.821688
etanni (click) 31710 27867 12 96 0 0% 0 0 3.231132
fannkuchredux (click) 24912 17701 9 257 0 0% 0 0 7.821632
fluentd (click) 389820 321093 11 111 0 0% 0 0 3.576317
graphql (click) 444390 375139 70 673 0 0% 0 0 20.750132
graphql-native (click) 409945 344835 41 271 0 0% 0 0 9.086465
lee (click) 316023 278067 49 731 0 0% 0 0 22.772743
matmul (click) 12586 12175 12 126 0 0% 0 0 4.178025
nbody (click) 15103 7574 11 185 0 0% 0 0 4.994059
nqueens (click) 23609 17866 10 261 0 0% 0 0 7.76966
optcarrot (click) 317087 278432 192 4342 34 0% 0 0 102.963876
rack (click) 272700 235097 31 322 0 0% 0 0 9.384104
ruby-json (click) 28135 23070 12 198 0 0% 0 0 5.770441
rubykon (click) 133625 106471 143 1498 3 0% 0 0 41.876459
sudoku (click) 47007 45195 11 507 0 0% 0 0 15.455754
tinygql (click) 313186 266445 63 788 5 0% 0 0 22.684351
30k_ifelse (click) 5222597 4374236 9264 50813 0 0% 0 0 1467.012957
30k_methods (click) 2019614 1594052 5782 19369 0 0% 0 0 518.257109
cfunc_itself (click) 7589 5810 9 70 0 0% 0 0 2.211014
fib (click) 5017 4633 8 49 0 0% 0 0 1.776456
getivar (click) 6102 6023 8 74 0 0% 0 0 2.117951
keyword_args (click) 8472 6502 10 72 0 0% 0 0 2.525021
respond_to (click) 8341 7760 9 85 0 0% 0 0 2.639201
ruby-xor (click) 7940 7126 9 87 0 0% 0 0 3.020671
setivar (click) 5193 4604 8 54 0 0% 0 0 1.734539
setivar_object (click) 5561 4624 8 54 0 0% 0 0 1.832523
setivar_young (click) 6284 5274 9 62 0 0% 0 0 2.014666
str_concat (click) 7932 7204 11 85 0 0% 0 0 2.888032
throw (click) 7946 6065 10 69 0 0% 0 0 2.448604

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.