YJIT Benchmarks

Details for Benchmarks at 2024-04-23 06:08:13 UTC

YJIT metrics from the yjit-bench suite using Ruby a53435868e

Using the geomean of the headline benchmarks for x86 YJIT 3.4.0dev is
  • 72.2% faster than CRuby 3.4.0dev
On railsbench it is
  • 79.2% 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 30 522 30 522
chunky-png 30 31 30 31
erubi-rails 30 19 30 19
hexapdf 30 15 30 15
liquid-c 30 345 30 345
liquid-compile 30 355 30 355
liquid-render 30 248 30 248
lobsters 30 23 30 23
mail 30 163 30 163
psych-load 30 15 30 15
railsbench 30 15 30 15
rubocop 30 183 30 183
ruby-lsp 30 175 30 175
sequel 30 289 30 289
binarytrees 30 86 30 86
blurhash 30 102 30 102
erubi 30 76 30 76
etanni 30 23 30 23
fannkuchredux 30 23 30 23
fluentd 30 15 30 15
graphql 30 15 30 15
graphql-native 30 32 30 32
lee 30 20 30 20
matmul 30 18 30 18
nbody 30 337 30 337
nqueens 30 293 30 293
optcarrot 30 15 30 15
protoboeuf 30 388 30 388
rack 30 180 30 180
ruby-json 30 15 30 15
rubykon 30 15 30 15
sudoku 30 25 30 25
tinygql 30 41 30 41
30k_ifelse 30 55 30 55
30k_methods 30 23 30 23
cfunc_itself 30 827 30 827
fib 30 417 30 417
getivar 30 942 30 942
keyword_args 30 591 30 591
respond_to 30 1480 30 1480
ruby-xor 30 97 30 97
setivar 30 2392 30 2392
setivar_object 30 384 30 384
setivar_young 30 412 30 412
str_concat 30 422 30 422
throw 30 765 30 765

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) 886524 770643 49 531 0 0% 0 0 16.098601
chunky-png (click) 337058 281965 87 1094 1 0% 0 0 32.926718
erubi-rails (click) 1709771 1439971 276 3004 6 0% 0 0 85.401903
hexapdf (click) 1416392 1231575 595 12688 38 0% 0 0 374.328616
liquid-c (click) 522579 453885 119 1757 5 0% 0 0 48.718836
liquid-compile (click) 446083 388339 151 2078 2 0% 0 0 63.772483
liquid-render (click) 622795 564427 136 2292 8 0% 0 0 63.226531
lobsters (click) 7873725 6569461 3212 51221 73 0% 0 0 1736.050511
mail (click) 730565 608610 346 4865 14 0% 0 0 143.546347
psych-load (click) 301118 244545 65 656 3 0% 0 0 20.696835
railsbench (click) 2995918 2594837 1634 15929 54 0% 0 0 459.346642
rubocop (click) 5369064 4579047 2901 48250 109 0% 4 0 1447.263276
ruby-lsp (click) 630782 554620 159 2794 14 0% 0 0 77.634115
sequel (click) 496800 419726 16 117 0 0% 0 0 4.259055
binarytrees (click) 9384 7556 11 75 0 0% 0 0 3.035826
blurhash (click) 51756 46977 32 460 0 0% 0 0 15.812624
erubi (click) 270262 237443 10 112 0 0% 0 0 3.900245
etanni (click) 29895 26052 11 91 0 0% 0 0 3.024369
fannkuchredux (click) 24443 9175 9 257 0 0% 0 0 8.315178
fluentd (click) 387439 335352 11 111 0 0% 0 0 3.676407
graphql (click) 440378 370375 70 685 0 0% 0 0 20.634507
graphql-native (click) 407894 342285 42 278 0 0% 0 0 9.250855
lee (click) 316265 260791 49 737 0 0% 0 0 23.375523
matmul (click) 12555 12107 12 126 0 0% 0 0 4.330938
nbody (click) 15072 7506 11 185 0 0% 0 0 5.192833
nqueens (click) 23358 17741 10 260 0 0% 0 0 8.486964
optcarrot (click) 313653 276483 192 4326 34 0% 0 0 105.173782
protoboeuf (click) 153148 132196 17 1378 0 0% 0 0 42.555094
rack (click) 273471 235907 31 345 0 0% 0 0 9.965098
ruby-json (click) 33192 28025 12 196 0 0% 0 0 5.945941
rubykon (click) 135398 124154 143 1517 3 0% 0 0 42.892098
sudoku (click) 46784 53075 11 507 0 0% 0 0 15.54149
tinygql (click) 313143 257403 63 787 5 0% 0 0 23.27202
30k_ifelse (click) 5222589 4374204 9264 50813 0 0% 0 0 1482.865609
30k_methods (click) 2019606 1594020 5782 19369 0 0% 0 0 528.853015
cfunc_itself (click) 7581 5778 9 70 0 0% 0 0 2.747918
fib (click) 5009 4601 8 49 0 0% 0 0 2.300696
getivar (click) 6094 5991 8 74 0 0% 0 0 2.58276
keyword_args (click) 8464 6470 10 72 0 0% 0 0 3.018407
respond_to (click) 8333 7728 9 85 0 0% 0 0 3.163649
ruby-xor (click) 8062 6902 9 87 0 0% 0 0 2.823346
setivar (click) 5185 4572 8 54 0 0% 0 0 2.252192
setivar_object (click) 5553 4592 8 54 0 0% 0 0 1.898471
setivar_young (click) 6276 5242 9 62 0 0% 0 0 2.102876
str_concat (click) 7876 7217 11 85 0 0% 0 0 2.946546
throw (click) 7938 6033 10 69 0 0% 0 0 2.431996

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.