finish iterator draft

This commit is contained in:
kageru 2020-04-12 14:34:52 +02:00
parent b4225d50e0
commit ba11ce21df
Signed by: kageru
GPG Key ID: 8282A2BEA4ADA3D2
3 changed files with 1273 additions and 785 deletions

@ -0,0 +1,933 @@
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Java.loop
# Run progress: 0.00% complete, ETA 01:31:40
# Fork: 1 of 5
# Warmup Iteration 1: 456.332 ms/op
# Warmup Iteration 2: 445.904 ms/op
# Warmup Iteration 3: 447.161 ms/op
# Warmup Iteration 4: 446.954 ms/op
# Warmup Iteration 5: 445.977 ms/op
Iteration 1: 446.109 ms/op
Iteration 2: 447.066 ms/op
Iteration 3: 445.489 ms/op
Iteration 4: 447.256 ms/op
Iteration 5: 447.188 ms/op
# Run progress: 1.82% complete, ETA 01:32:24
# Fork: 2 of 5
# Warmup Iteration 1: 452.201 ms/op
# Warmup Iteration 2: 445.484 ms/op
# Warmup Iteration 3: 447.031 ms/op
# Warmup Iteration 4: 445.762 ms/op
# Warmup Iteration 5: 445.223 ms/op
Iteration 1: 446.364 ms/op
Iteration 2: 445.127 ms/op
Iteration 3: 445.872 ms/op
Iteration 4: 447.157 ms/op
Iteration 5: 445.903 ms/op
# Run progress: 3.64% complete, ETA 01:30:48
# Fork: 3 of 5
# Warmup Iteration 1: 450.744 ms/op
# Warmup Iteration 2: 446.860 ms/op
# Warmup Iteration 3: 447.187 ms/op
# Warmup Iteration 4: 445.894 ms/op
# Warmup Iteration 5: 446.363 ms/op
Iteration 1: 446.135 ms/op
Iteration 2: 445.913 ms/op
Iteration 3: 445.983 ms/op
Iteration 4: 445.214 ms/op
Iteration 5: 445.505 ms/op
# Run progress: 5.45% complete, ETA 01:29:07
# Fork: 4 of 5
# Warmup Iteration 1: 453.909 ms/op
# Warmup Iteration 2: 445.268 ms/op
# Warmup Iteration 3: 447.106 ms/op
# Warmup Iteration 4: 452.267 ms/op
# Warmup Iteration 5: 446.321 ms/op
Iteration 1: 447.768 ms/op
Iteration 2: 446.519 ms/op
Iteration 3: 445.002 ms/op
Iteration 4: 447.406 ms/op
Iteration 5: 444.889 ms/op
# Run progress: 7.27% complete, ETA 01:27:28
# Fork: 5 of 5
# Warmup Iteration 1: 453.557 ms/op
# Warmup Iteration 2: 446.911 ms/op
# Warmup Iteration 3: 447.795 ms/op
# Warmup Iteration 4: 445.091 ms/op
# Warmup Iteration 5: 447.698 ms/op
Iteration 1: 444.943 ms/op
Iteration 2: 445.156 ms/op
Iteration 3: 447.101 ms/op
Iteration 4: 445.289 ms/op
Iteration 5: 445.011 ms/op
Result "perf.Java.loop":
446.055 ±(99.9%) 0.677 ms/op [Average]
(min, avg, max) = (444.889, 446.055, 447.768), stdev = 0.904
CI (99.9%): [445.378, 446.732] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Java.stream
# Run progress: 9.09% complete, ETA 01:25:45
# Fork: 1 of 5
# Warmup Iteration 1: 683.149 ms/op
# Warmup Iteration 2: 591.776 ms/op
# Warmup Iteration 3: 591.627 ms/op
# Warmup Iteration 4: 598.434 ms/op
# Warmup Iteration 5: 602.405 ms/op
Iteration 1: 632.719 ms/op
Iteration 2: 594.098 ms/op
Iteration 3: 591.569 ms/op
Iteration 4: 592.402 ms/op
Iteration 5: 595.107 ms/op
# Run progress: 10.91% complete, ETA 01:23:50
# Fork: 2 of 5
# Warmup Iteration 1: 678.576 ms/op
# Warmup Iteration 2: 591.063 ms/op
# Warmup Iteration 3: 594.322 ms/op
# Warmup Iteration 4: 591.848 ms/op
# Warmup Iteration 5: 594.417 ms/op
Iteration 1: 595.271 ms/op
Iteration 2: 592.028 ms/op
Iteration 3: 642.472 ms/op
Iteration 4: 593.947 ms/op
Iteration 5: 592.536 ms/op
# Run progress: 12.73% complete, ETA 01:21:58
# Fork: 3 of 5
# Warmup Iteration 1: 678.283 ms/op
# Warmup Iteration 2: 594.497 ms/op
# Warmup Iteration 3: 592.839 ms/op
# Warmup Iteration 4: 592.898 ms/op
# Warmup Iteration 5: 594.148 ms/op
Iteration 1: 629.247 ms/op
Iteration 2: 592.774 ms/op
Iteration 3: 593.806 ms/op
Iteration 4: 593.348 ms/op
Iteration 5: 593.438 ms/op
# Run progress: 14.55% complete, ETA 01:20:08
# Fork: 4 of 5
# Warmup Iteration 1: 679.167 ms/op
# Warmup Iteration 2: 591.866 ms/op
# Warmup Iteration 3: 592.856 ms/op
# Warmup Iteration 4: 592.900 ms/op
# Warmup Iteration 5: 591.494 ms/op
Iteration 1: 635.757 ms/op
Iteration 2: 593.654 ms/op
Iteration 3: 591.762 ms/op
Iteration 4: 593.685 ms/op
Iteration 5: 592.605 ms/op
# Run progress: 16.36% complete, ETA 01:18:20
# Fork: 5 of 5
# Warmup Iteration 1: 679.848 ms/op
# Warmup Iteration 2: 594.060 ms/op
# Warmup Iteration 3: 592.618 ms/op
# Warmup Iteration 4: 596.782 ms/op
# Warmup Iteration 5: 593.338 ms/op
Iteration 1: 630.190 ms/op
Iteration 2: 594.034 ms/op
Iteration 3: 594.066 ms/op
Iteration 4: 592.662 ms/op
Iteration 5: 592.430 ms/op
Result "perf.Java.stream":
601.424 ±(99.9%) 12.606 ms/op [Average]
(min, avg, max) = (591.569, 601.424, 642.472), stdev = 16.829
CI (99.9%): [588.818, 614.031] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.loop
# Run progress: 18.18% complete, ETA 01:16:33
# Fork: 1 of 5
# Warmup Iteration 1: 452.772 ms/op
# Warmup Iteration 2: 445.813 ms/op
# Warmup Iteration 3: 445.203 ms/op
# Warmup Iteration 4: 447.020 ms/op
# Warmup Iteration 5: 445.761 ms/op
Iteration 1: 445.277 ms/op
Iteration 2: 447.734 ms/op
Iteration 3: 445.433 ms/op
Iteration 4: 445.653 ms/op
Iteration 5: 446.958 ms/op
# Run progress: 20.00% complete, ETA 01:14:54
# Fork: 2 of 5
# Warmup Iteration 1: 450.903 ms/op
# Warmup Iteration 2: 445.990 ms/op
# Warmup Iteration 3: 447.923 ms/op
# Warmup Iteration 4: 445.835 ms/op
# Warmup Iteration 5: 445.211 ms/op
Iteration 1: 447.195 ms/op
Iteration 2: 449.569 ms/op
Iteration 3: 444.934 ms/op
Iteration 4: 446.823 ms/op
Iteration 5: 445.366 ms/op
# Run progress: 21.82% complete, ETA 01:13:15
# Fork: 3 of 5
# Warmup Iteration 1: 456.334 ms/op
# Warmup Iteration 2: 446.712 ms/op
# Warmup Iteration 3: 445.541 ms/op
# Warmup Iteration 4: 446.043 ms/op
# Warmup Iteration 5: 446.722 ms/op
Iteration 1: 445.405 ms/op
Iteration 2: 447.530 ms/op
Iteration 3: 450.772 ms/op
Iteration 4: 445.377 ms/op
Iteration 5: 447.160 ms/op
# Run progress: 23.64% complete, ETA 01:11:34
# Fork: 4 of 5
# Warmup Iteration 1: 452.581 ms/op
# Warmup Iteration 2: 445.681 ms/op
# Warmup Iteration 3: 447.341 ms/op
# Warmup Iteration 4: 446.772 ms/op
# Warmup Iteration 5: 447.728 ms/op
Iteration 1: 447.228 ms/op
Iteration 2: 446.791 ms/op
Iteration 3: 445.374 ms/op
Iteration 4: 447.020 ms/op
Iteration 5: 445.255 ms/op
# Run progress: 25.45% complete, ETA 01:09:54
# Fork: 5 of 5
# Warmup Iteration 1: 450.730 ms/op
# Warmup Iteration 2: 447.652 ms/op
# Warmup Iteration 3: 445.747 ms/op
# Warmup Iteration 4: 444.973 ms/op
# Warmup Iteration 5: 446.593 ms/op
Iteration 1: 445.315 ms/op
Iteration 2: 445.712 ms/op
Iteration 3: 449.632 ms/op
Iteration 4: 446.242 ms/op
Iteration 5: 445.253 ms/op
Result "perf.Kotlin.loop":
446.600 ±(99.9%) 1.164 ms/op [Average]
(min, avg, max) = (444.934, 446.600, 450.772), stdev = 1.554
CI (99.9%): [445.436, 447.764] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.sequence
# Run progress: 27.27% complete, ETA 01:08:14
# Fork: 1 of 5
# Warmup Iteration 1: 2875.137 ms/op
# Warmup Iteration 2: 2714.992 ms/op
# Warmup Iteration 3: 2717.500 ms/op
# Warmup Iteration 4: 2722.842 ms/op
# Warmup Iteration 5: 2726.368 ms/op
Iteration 1: 2718.843 ms/op
Iteration 2: 2730.963 ms/op
Iteration 3: 2725.803 ms/op
Iteration 4: 2735.491 ms/op
Iteration 5: 2720.835 ms/op
# Run progress: 29.09% complete, ETA 01:06:49
# Fork: 2 of 5
# Warmup Iteration 1: 2872.714 ms/op
# Warmup Iteration 2: 2737.939 ms/op
# Warmup Iteration 3: 2723.591 ms/op
# Warmup Iteration 4: 2739.880 ms/op
# Warmup Iteration 5: 2735.399 ms/op
Iteration 1: 2725.004 ms/op
Iteration 2: 2739.529 ms/op
Iteration 3: 2735.921 ms/op
Iteration 4: 2737.889 ms/op
Iteration 5: 2745.381 ms/op
# Run progress: 30.91% complete, ETA 01:05:23
# Fork: 3 of 5
# Warmup Iteration 1: 2863.006 ms/op
# Warmup Iteration 2: 2726.449 ms/op
# Warmup Iteration 3: 2733.619 ms/op
# Warmup Iteration 4: 2725.596 ms/op
# Warmup Iteration 5: 2743.223 ms/op
Iteration 1: 2730.597 ms/op
Iteration 2: 2742.830 ms/op
Iteration 3: 2737.714 ms/op
Iteration 4: 2727.641 ms/op
Iteration 5: 2729.501 ms/op
# Run progress: 32.73% complete, ETA 01:03:54
# Fork: 4 of 5
# Warmup Iteration 1: 2874.924 ms/op
# Warmup Iteration 2: 2727.463 ms/op
# Warmup Iteration 3: 2792.965 ms/op
# Warmup Iteration 4: 2722.637 ms/op
# Warmup Iteration 5: 2724.197 ms/op
Iteration 1: 2736.731 ms/op
Iteration 2: 2726.640 ms/op
Iteration 3: 2748.462 ms/op
Iteration 4: 2747.123 ms/op
Iteration 5: 2741.907 ms/op
# Run progress: 34.55% complete, ETA 01:02:23
# Fork: 5 of 5
# Warmup Iteration 1: 2852.575 ms/op
# Warmup Iteration 2: 2727.111 ms/op
# Warmup Iteration 3: 2734.597 ms/op
# Warmup Iteration 4: 2731.829 ms/op
# Warmup Iteration 5: 2720.765 ms/op
Iteration 1: 2724.821 ms/op
Iteration 2: 2733.517 ms/op
Iteration 3: 2726.031 ms/op
Iteration 4: 2714.774 ms/op
Iteration 5: 2731.156 ms/op
Result "perf.Kotlin.sequence":
2732.604 ±(99.9%) 6.644 ms/op [Average]
(min, avg, max) = (2714.774, 2732.604, 2748.462), stdev = 8.870
CI (99.9%): [2725.960, 2739.248] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.stream
# Run progress: 36.36% complete, ETA 01:00:49
# Fork: 1 of 5
# Warmup Iteration 1: 678.345 ms/op
# Warmup Iteration 2: 591.611 ms/op
# Warmup Iteration 3: 593.355 ms/op
# Warmup Iteration 4: 593.125 ms/op
# Warmup Iteration 5: 592.035 ms/op
Iteration 1: 594.764 ms/op
Iteration 2: 595.813 ms/op
Iteration 3: 592.359 ms/op
Iteration 4: 593.853 ms/op
Iteration 5: 593.184 ms/op
# Run progress: 38.18% complete, ETA 00:59:00
# Fork: 2 of 5
# Warmup Iteration 1: 680.533 ms/op
# Warmup Iteration 2: 644.939 ms/op
# Warmup Iteration 3: 592.135 ms/op
# Warmup Iteration 4: 592.386 ms/op
# Warmup Iteration 5: 593.396 ms/op
Iteration 1: 592.626 ms/op
Iteration 2: 593.254 ms/op
Iteration 3: 594.582 ms/op
Iteration 4: 591.622 ms/op
Iteration 5: 592.595 ms/op
# Run progress: 40.00% complete, ETA 00:57:11
# Fork: 3 of 5
# Warmup Iteration 1: 680.298 ms/op
# Warmup Iteration 2: 591.609 ms/op
# Warmup Iteration 3: 592.565 ms/op
# Warmup Iteration 4: 591.873 ms/op
# Warmup Iteration 5: 591.836 ms/op
Iteration 1: 593.007 ms/op
Iteration 2: 592.916 ms/op
Iteration 3: 591.434 ms/op
Iteration 4: 594.996 ms/op
Iteration 5: 592.853 ms/op
# Run progress: 41.82% complete, ETA 00:55:23
# Fork: 4 of 5
# Warmup Iteration 1: 678.935 ms/op
# Warmup Iteration 2: 595.014 ms/op
# Warmup Iteration 3: 593.754 ms/op
# Warmup Iteration 4: 592.117 ms/op
# Warmup Iteration 5: 593.827 ms/op
Iteration 1: 591.940 ms/op
Iteration 2: 591.284 ms/op
Iteration 3: 593.308 ms/op
Iteration 4: 597.758 ms/op
Iteration 5: 591.519 ms/op
# Run progress: 43.64% complete, ETA 00:53:36
# Fork: 5 of 5
# Warmup Iteration 1: 681.321 ms/op
# Warmup Iteration 2: 635.465 ms/op
# Warmup Iteration 3: 591.332 ms/op
# Warmup Iteration 4: 593.023 ms/op
# Warmup Iteration 5: 590.999 ms/op
Iteration 1: 590.586 ms/op
Iteration 2: 597.205 ms/op
Iteration 3: 591.966 ms/op
Iteration 4: 592.190 ms/op
Iteration 5: 596.203 ms/op
Result "perf.Kotlin.stream":
593.353 ±(99.9%) 1.408 ms/op [Average]
(min, avg, max) = (590.586, 593.353, 597.758), stdev = 1.880
CI (99.9%): [591.945, 594.761] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.streamWrappedInSequence
# Run progress: 45.45% complete, ETA 00:51:49
# Fork: 1 of 5
# Warmup Iteration 1: 3989.143 ms/op
# Warmup Iteration 2: 3808.328 ms/op
# Warmup Iteration 3: 3851.896 ms/op
# Warmup Iteration 4: 3821.087 ms/op
# Warmup Iteration 5: 3842.988 ms/op
Iteration 1: 3838.952 ms/op
Iteration 2: 3828.666 ms/op
Iteration 3: 3829.225 ms/op
Iteration 4: 3804.143 ms/op
Iteration 5: 3803.303 ms/op
# Run progress: 47.27% complete, ETA 00:50:19
# Fork: 2 of 5
# Warmup Iteration 1: 3965.935 ms/op
# Warmup Iteration 2: 3830.953 ms/op
# Warmup Iteration 3: 3831.254 ms/op
# Warmup Iteration 4: 3786.462 ms/op
# Warmup Iteration 5: 3810.373 ms/op
Iteration 1: 3820.851 ms/op
Iteration 2: 3850.427 ms/op
Iteration 3: 3874.084 ms/op
Iteration 4: 3810.283 ms/op
Iteration 5: 3794.962 ms/op
# Run progress: 49.09% complete, ETA 00:48:46
# Fork: 3 of 5
# Warmup Iteration 1: 3971.302 ms/op
# Warmup Iteration 2: 3805.030 ms/op
# Warmup Iteration 3: 3827.273 ms/op
# Warmup Iteration 4: 3797.691 ms/op
# Warmup Iteration 5: 3828.704 ms/op
Iteration 1: 3835.989 ms/op
Iteration 2: 3807.760 ms/op
Iteration 3: 4022.427 ms/op
Iteration 4: 3809.359 ms/op
Iteration 5: 3834.482 ms/op
# Run progress: 50.91% complete, ETA 00:47:12
# Fork: 4 of 5
# Warmup Iteration 1: 4038.999 ms/op
# Warmup Iteration 2: 3790.741 ms/op
# Warmup Iteration 3: 3816.597 ms/op
# Warmup Iteration 4: 3804.865 ms/op
# Warmup Iteration 5: 3835.732 ms/op
Iteration 1: 3804.229 ms/op
Iteration 2: 3854.098 ms/op
Iteration 3: 3822.642 ms/op
Iteration 4: 3817.002 ms/op
Iteration 5: 3804.801 ms/op
# Run progress: 52.73% complete, ETA 00:45:37
# Fork: 5 of 5
# Warmup Iteration 1: 3975.357 ms/op
# Warmup Iteration 2: 3848.500 ms/op
# Warmup Iteration 3: 3794.490 ms/op
# Warmup Iteration 4: 3823.770 ms/op
# Warmup Iteration 5: 3810.056 ms/op
Iteration 1: 3831.363 ms/op
Iteration 2: 3819.838 ms/op
Iteration 3: 3787.684 ms/op
Iteration 4: 3808.174 ms/op
Iteration 5: 3815.488 ms/op
Result "perf.Kotlin.streamWrappedInSequence":
3829.209 ±(99.9%) 33.569 ms/op [Average]
(min, avg, max) = (3787.684, 3829.209, 4022.427), stdev = 44.814
CI (99.9%): [3795.640, 3862.779] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.withGenerator
# Run progress: 54.55% complete, ETA 00:44:00
# Fork: 1 of 5
# Warmup Iteration 1: 10059.659 ms/op
# Warmup Iteration 2: 8877.213 ms/op
# Warmup Iteration 3: 8124.623 ms/op
# Warmup Iteration 4: 8131.638 ms/op
# Warmup Iteration 5: 8148.939 ms/op
Iteration 1: 8148.879 ms/op
Iteration 2: 8055.587 ms/op
Iteration 3: 8173.149 ms/op
Iteration 4: 8169.328 ms/op
Iteration 5: 8132.040 ms/op
# Run progress: 56.36% complete, ETA 00:42:55
# Fork: 2 of 5
# Warmup Iteration 1: 10101.875 ms/op
# Warmup Iteration 2: 8876.112 ms/op
# Warmup Iteration 3: 8091.701 ms/op
# Warmup Iteration 4: 8130.830 ms/op
# Warmup Iteration 5: 8162.659 ms/op
Iteration 1: 8112.434 ms/op
Iteration 2: 8077.144 ms/op
Iteration 3: 8134.760 ms/op
Iteration 4: 8178.659 ms/op
Iteration 5: 8122.861 ms/op
# Run progress: 58.18% complete, ETA 00:41:44
# Fork: 3 of 5
# Warmup Iteration 1: 10088.192 ms/op
# Warmup Iteration 2: 8938.462 ms/op
# Warmup Iteration 3: 8169.118 ms/op
# Warmup Iteration 4: 8110.177 ms/op
# Warmup Iteration 5: 8120.017 ms/op
Iteration 1: 8130.142 ms/op
Iteration 2: 8146.593 ms/op
Iteration 3: 8161.427 ms/op
Iteration 4: 8180.565 ms/op
Iteration 5: 8153.313 ms/op
# Run progress: 60.00% complete, ETA 00:40:28
# Fork: 4 of 5
# Warmup Iteration 1: 12624.377 ms/op
# Warmup Iteration 2: 8872.746 ms/op
# Warmup Iteration 3: 8114.339 ms/op
# Warmup Iteration 4: 8144.090 ms/op
# Warmup Iteration 5: 8054.094 ms/op
Iteration 1: 8134.096 ms/op
Iteration 2: 8154.380 ms/op
Iteration 3: 8130.686 ms/op
Iteration 4: 14015.371 ms/op
Iteration 5: 8114.715 ms/op
# Run progress: 61.82% complete, ETA 00:39:07
# Fork: 5 of 5
# Warmup Iteration 1: 10102.610 ms/op
# Warmup Iteration 2: 8880.406 ms/op
# Warmup Iteration 3: 8116.952 ms/op
# Warmup Iteration 4: 8223.242 ms/op
# Warmup Iteration 5: 8149.098 ms/op
Iteration 1: 8152.230 ms/op
Iteration 2: 8110.804 ms/op
Iteration 3: 8136.879 ms/op
Iteration 4: 8178.455 ms/op
Iteration 5: 8149.239 ms/op
Result "perf.Kotlin.withGenerator":
8374.149 ±(99.9%) 880.647 ms/op [Average]
(min, avg, max) = (8055.587, 8374.149, 14015.371), stdev = 1175.639
CI (99.9%): [7493.502, 9254.797] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.loop
# Run progress: 63.64% complete, ETA 00:37:42
# Fork: 1 of 5
# Warmup Iteration 1: 454.126 ms/op
# Warmup Iteration 2: 446.612 ms/op
# Warmup Iteration 3: 445.577 ms/op
# Warmup Iteration 4: 447.197 ms/op
# Warmup Iteration 5: 445.174 ms/op
Iteration 1: 444.831 ms/op
Iteration 2: 446.933 ms/op
Iteration 3: 445.256 ms/op
Iteration 4: 444.518 ms/op
Iteration 5: 447.351 ms/op
# Run progress: 65.45% complete, ETA 00:35:44
# Fork: 2 of 5
# Warmup Iteration 1: 453.077 ms/op
# Warmup Iteration 2: 445.972 ms/op
# Warmup Iteration 3: 448.129 ms/op
# Warmup Iteration 4: 446.268 ms/op
# Warmup Iteration 5: 444.718 ms/op
Iteration 1: 447.130 ms/op
Iteration 2: 445.217 ms/op
Iteration 3: 445.393 ms/op
Iteration 4: 447.177 ms/op
Iteration 5: 445.732 ms/op
# Run progress: 67.27% complete, ETA 00:33:46
# Fork: 3 of 5
# Warmup Iteration 1: 451.974 ms/op
# Warmup Iteration 2: 446.601 ms/op
# Warmup Iteration 3: 445.653 ms/op
# Warmup Iteration 4: 448.023 ms/op
# Warmup Iteration 5: 458.473 ms/op
Iteration 1: 447.401 ms/op
Iteration 2: 445.530 ms/op
Iteration 3: 446.670 ms/op
Iteration 4: 444.751 ms/op
Iteration 5: 446.407 ms/op
# Run progress: 69.09% complete, ETA 00:31:49
# Fork: 4 of 5
# Warmup Iteration 1: 452.980 ms/op
# Warmup Iteration 2: 444.762 ms/op
# Warmup Iteration 3: 447.634 ms/op
# Warmup Iteration 4: 446.151 ms/op
# Warmup Iteration 5: 444.367 ms/op
Iteration 1: 446.652 ms/op
Iteration 2: 445.799 ms/op
Iteration 3: 445.204 ms/op
Iteration 4: 446.163 ms/op
Iteration 5: 446.287 ms/op
# Run progress: 70.91% complete, ETA 00:29:53
# Fork: 5 of 5
# Warmup Iteration 1: 451.592 ms/op
# Warmup Iteration 2: 448.889 ms/op
# Warmup Iteration 3: 445.963 ms/op
# Warmup Iteration 4: 445.615 ms/op
# Warmup Iteration 5: 447.028 ms/op
Iteration 1: 445.457 ms/op
Iteration 2: 445.110 ms/op
Iteration 3: 446.664 ms/op
Iteration 4: 445.987 ms/op
Iteration 5: 446.179 ms/op
Result "perf.NonPrimitive.loop":
445.992 ±(99.9%) 0.642 ms/op [Average]
(min, avg, max) = (444.518, 445.992, 447.401), stdev = 0.858
CI (99.9%): [445.349, 446.634] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.parallelStream
# Run progress: 72.73% complete, ETA 00:27:57
# Fork: 1 of 5
# Warmup Iteration 1: 31838.855 ms/op
# Warmup Iteration 2: 34561.656 ms/op
# Warmup Iteration 3: 35669.703 ms/op
# Warmup Iteration 4: 34349.624 ms/op
# Warmup Iteration 5: 33726.583 ms/op
Iteration 1: 33877.354 ms/op
Iteration 2: 33822.699 ms/op
Iteration 3: 33778.386 ms/op
Iteration 4: 34008.675 ms/op
Iteration 5: 33856.945 ms/op
# Run progress: 74.55% complete, ETA 00:27:23
# Fork: 2 of 5
# Warmup Iteration 1: 32494.459 ms/op
# Warmup Iteration 2: 36153.069 ms/op
# Warmup Iteration 3: 36796.548 ms/op
# Warmup Iteration 4: 35025.518 ms/op
# Warmup Iteration 5: 33258.558 ms/op
Iteration 1: 33554.978 ms/op
Iteration 2: 33424.136 ms/op
Iteration 3: 33627.128 ms/op
Iteration 4: 34313.959 ms/op
Iteration 5: 34112.359 ms/op
# Run progress: 76.36% complete, ETA 00:26:36
# Fork: 3 of 5
# Warmup Iteration 1: 31567.609 ms/op
# Warmup Iteration 2: 34712.718 ms/op
# Warmup Iteration 3: 33958.542 ms/op
# Warmup Iteration 4: 34035.515 ms/op
# Warmup Iteration 5: 34438.473 ms/op
Iteration 1: 33670.314 ms/op
Iteration 2: 33864.096 ms/op
Iteration 3: 34302.070 ms/op
Iteration 4: 33971.974 ms/op
Iteration 5: 33560.701 ms/op
# Run progress: 78.18% complete, ETA 00:25:33
# Fork: 4 of 5
# Warmup Iteration 1: 31742.222 ms/op
# Warmup Iteration 2: 34731.781 ms/op
# Warmup Iteration 3: 33450.090 ms/op
# Warmup Iteration 4: 33130.877 ms/op
# Warmup Iteration 5: 33546.489 ms/op
Iteration 1: 33366.136 ms/op
Iteration 2: 33895.648 ms/op
Iteration 3: 34376.907 ms/op
Iteration 4: 33857.023 ms/op
Iteration 5: 33111.330 ms/op
# Run progress: 80.00% complete, ETA 00:24:17
# Fork: 5 of 5
# Warmup Iteration 1: 31518.688 ms/op
# Warmup Iteration 2: 34623.682 ms/op
# Warmup Iteration 3: 34725.832 ms/op
# Warmup Iteration 4: 33545.596 ms/op
# Warmup Iteration 5: 34196.732 ms/op
Iteration 1: 34093.607 ms/op
Iteration 2: 33647.008 ms/op
Iteration 3: 34429.111 ms/op
Iteration 4: 33643.974 ms/op
Iteration 5: 34256.452 ms/op
Result "perf.NonPrimitive.parallelStream":
33856.919 ±(99.9%) 249.911 ms/op [Average]
(min, avg, max) = (33111.330, 33856.919, 34429.111), stdev = 333.625
CI (99.9%): [33607.007, 34106.830] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.sequence
# Run progress: 81.82% complete, ETA 00:22:51
# Fork: 1 of 5
# Warmup Iteration 1: 29376.336 ms/op
# Warmup Iteration 2: 27208.353 ms/op
# Warmup Iteration 3: 27221.251 ms/op
# Warmup Iteration 4: 27217.165 ms/op
# Warmup Iteration 5: 27146.583 ms/op
Iteration 1: 27302.438 ms/op
Iteration 2: 27158.834 ms/op
Iteration 3: 29219.908 ms/op
Iteration 4: 27236.755 ms/op
Iteration 5: 27213.294 ms/op
# Run progress: 83.64% complete, ETA 00:21:01
# Fork: 2 of 5
# Warmup Iteration 1: 27179.841 ms/op
# Warmup Iteration 2: 27104.263 ms/op
# Warmup Iteration 3: 27025.071 ms/op
# Warmup Iteration 4: 27029.083 ms/op
# Warmup Iteration 5: 27048.692 ms/op
Iteration 1: 27151.655 ms/op
Iteration 2: 27469.413 ms/op
Iteration 3: 26944.107 ms/op
Iteration 4: 27028.756 ms/op
Iteration 5: 27077.869 ms/op
# Run progress: 85.45% complete, ETA 00:19:03
# Fork: 3 of 5
# Warmup Iteration 1: 27227.611 ms/op
# Warmup Iteration 2: 27068.450 ms/op
# Warmup Iteration 3: 27032.028 ms/op
# Warmup Iteration 4: 28617.949 ms/op
# Warmup Iteration 5: 26948.167 ms/op
Iteration 1: 27025.160 ms/op
Iteration 2: 27070.999 ms/op
Iteration 3: 27105.065 ms/op
Iteration 4: 27025.640 ms/op
Iteration 5: 26987.588 ms/op
# Run progress: 87.27% complete, ETA 00:16:59
# Fork: 4 of 5
# Warmup Iteration 1: 27227.761 ms/op
# Warmup Iteration 2: 27007.131 ms/op
# Warmup Iteration 3: 27072.111 ms/op
# Warmup Iteration 4: 27429.696 ms/op
# Warmup Iteration 5: 27052.873 ms/op
Iteration 1: 27165.902 ms/op
Iteration 2: 27522.829 ms/op
Iteration 3: 27791.518 ms/op
Iteration 4: 27459.632 ms/op
Iteration 5: 27263.962 ms/op
# Run progress: 89.09% complete, ETA 00:14:49
# Fork: 5 of 5
# Warmup Iteration 1: 27245.157 ms/op
# Warmup Iteration 2: 27099.383 ms/op
# Warmup Iteration 3: 27043.463 ms/op
# Warmup Iteration 4: 26944.143 ms/op
# Warmup Iteration 5: 27185.798 ms/op
Iteration 1: 27204.744 ms/op
Iteration 2: 26961.133 ms/op
Iteration 3: 26975.095 ms/op
Iteration 4: 26976.901 ms/op
Iteration 5: 27095.789 ms/op
Result "perf.NonPrimitive.sequence":
27257.399 ±(99.9%) 342.686 ms/op [Average]
(min, avg, max) = (26944.107, 27257.399, 29219.908), stdev = 457.476
CI (99.9%): [26914.713, 27600.086] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.stream
# Run progress: 90.91% complete, ETA 00:12:33
# Fork: 1 of 5
# Warmup Iteration 1: 37907.647 ms/op
# Warmup Iteration 2: 44312.088 ms/op
# Warmup Iteration 3: 44351.894 ms/op
# Warmup Iteration 4: 45600.829 ms/op
# Warmup Iteration 5: 44292.840 ms/op
Iteration 1: 48919.294 ms/op
Iteration 2: 50229.844 ms/op
Iteration 3: 46366.206 ms/op
Iteration 4: 46053.252 ms/op
Iteration 5: 44309.572 ms/op
# Run progress: 92.73% complete, ETA 00:10:26
# Fork: 2 of 5
# Warmup Iteration 1: 36888.253 ms/op
# Warmup Iteration 2: 44609.411 ms/op
# Warmup Iteration 3: 43528.954 ms/op
# Warmup Iteration 4: 44225.429 ms/op
# Warmup Iteration 5: 43411.202 ms/op
Iteration 1: 43504.814 ms/op
Iteration 2: 43979.944 ms/op
Iteration 3: 43733.566 ms/op
Iteration 4: 43561.106 ms/op
Iteration 5: 44148.817 ms/op
# Run progress: 94.55% complete, ETA 00:08:05
# Fork: 3 of 5
# Warmup Iteration 1: 44621.184 ms/op
# Warmup Iteration 2: 43440.425 ms/op
# Warmup Iteration 3: 43381.843 ms/op
# Warmup Iteration 4: 43383.749 ms/op
# Warmup Iteration 5: 43329.551 ms/op
Iteration 1: 43357.752 ms/op
Iteration 2: 43427.313 ms/op
Iteration 3: 43366.149 ms/op
Iteration 4: 44638.673 ms/op
Iteration 5: 43953.000 ms/op
# Run progress: 96.36% complete, ETA 00:05:34
# Fork: 4 of 5
# Warmup Iteration 1: 37357.182 ms/op
# Warmup Iteration 2: 43539.552 ms/op
# Warmup Iteration 3: 43484.783 ms/op
# Warmup Iteration 4: 44086.420 ms/op
# Warmup Iteration 5: 43492.440 ms/op
Iteration 1: 47195.842 ms/op
Iteration 2: 43471.970 ms/op
Iteration 3: 43533.048 ms/op
Iteration 4: 44247.170 ms/op
Iteration 5: 43474.395 ms/op
# Run progress: 98.18% complete, ETA 00:02:52
# Fork: 5 of 5
# Warmup Iteration 1: 46786.069 ms/op
# Warmup Iteration 2: 44113.438 ms/op
# Warmup Iteration 3: 44319.157 ms/op
# Warmup Iteration 4: 44343.082 ms/op
# Warmup Iteration 5: 44142.656 ms/op
Iteration 1: 44516.563 ms/op
Iteration 2: 44090.180 ms/op
Iteration 3: 44287.859 ms/op
Iteration 4: 44238.998 ms/op
Iteration 5: 44227.628 ms/op
Result "perf.NonPrimitive.stream":
44673.318 ±(99.9%) 1325.832 ms/op [Average]
(min, avg, max) = (43357.752, 44673.318, 50229.844), stdev = 1769.948
CI (99.9%): [43347.486, 45999.150] (assumes normal distribution)
# Run complete. Total time: 02:42:16
REMEMBER: The numbers below are just data. To gain reusable insights, you need to follow up on
why the numbers are the way they are. Use profilers (see -prof, -lprof), design factorial
experiments, perform baseline and negative tests that provide experimental control, make sure
the benchmarking environment is safe on JVM/OS/HW level, ask for reviews from the domain experts.
Do not assume the numbers tell you what you want them to tell.
Benchmark Mode Cnt Score Error Units
Java.loop avgt 25 446.055 ± 0.677 ms/op
Java.stream avgt 25 601.424 ± 12.606 ms/op
Kotlin.loop avgt 25 446.600 ± 1.164 ms/op
Kotlin.sequence avgt 25 2732.604 ± 6.644 ms/op
Kotlin.stream avgt 25 593.353 ± 1.408 ms/op
Kotlin.streamWrappedInSequence avgt 25 3829.209 ± 33.569 ms/op
Kotlin.withGenerator avgt 25 8374.149 ± 880.647 ms/op
NonPrimitive.loop avgt 25 445.992 ± 0.642 ms/op
NonPrimitive.parallelStream avgt 25 33856.919 ± 249.911 ms/op
NonPrimitive.sequence avgt 25 27257.399 ± 342.686 ms/op
NonPrimitive.stream avgt 25 44673.318 ± 1325.832 ms/op

@ -1,765 +0,0 @@
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Java.loop
# Run progress: 0.00% complete, ETA 01:23:20
# Fork: 1 of 5
# Warmup Iteration 1: 503.008 ms/op
# Warmup Iteration 2: 497.339 ms/op
# Warmup Iteration 3: 497.504 ms/op
# Warmup Iteration 4: 526.782 ms/op
# Warmup Iteration 5: 515.432 ms/op
Iteration 1: 501.527 ms/op
Iteration 2: 499.508 ms/op
Iteration 3: 498.462 ms/op
Iteration 4: 492.443 ms/op
Iteration 5: 497.703 ms/op
# Run progress: 2.00% complete, ETA 01:24:21
# Fork: 2 of 5
# Warmup Iteration 1: 498.714 ms/op
# Warmup Iteration 2: 494.639 ms/op
# Warmup Iteration 3: 494.410 ms/op
# Warmup Iteration 4: 496.832 ms/op
# Warmup Iteration 5: 528.760 ms/op
Iteration 1: 495.342 ms/op
Iteration 2: 498.370 ms/op
Iteration 3: 495.366 ms/op
Iteration 4: 496.939 ms/op
Iteration 5: 496.600 ms/op
# Run progress: 4.00% complete, ETA 01:22:57
# Fork: 3 of 5
# Warmup Iteration 1: 510.115 ms/op
# Warmup Iteration 2: 502.292 ms/op
# Warmup Iteration 3: 503.220 ms/op
# Warmup Iteration 4: 500.617 ms/op
# Warmup Iteration 5: 514.586 ms/op
Iteration 1: 511.165 ms/op
Iteration 2: 494.087 ms/op
Iteration 3: 494.140 ms/op
Iteration 4: 494.629 ms/op
Iteration 5: 497.200 ms/op
# Run progress: 6.00% complete, ETA 01:20:56
# Fork: 4 of 5
# Warmup Iteration 1: 513.161 ms/op
# Warmup Iteration 2: 495.656 ms/op
# Warmup Iteration 3: 494.638 ms/op
# Warmup Iteration 4: 490.464 ms/op
# Warmup Iteration 5: 500.923 ms/op
Iteration 1: 496.505 ms/op
Iteration 2: 493.417 ms/op
Iteration 3: 497.258 ms/op
Iteration 4: 503.422 ms/op
Iteration 5: 496.757 ms/op
# Run progress: 8.00% complete, ETA 01:19:13
# Fork: 5 of 5
# Warmup Iteration 1: 508.611 ms/op
# Warmup Iteration 2: 497.342 ms/op
# Warmup Iteration 3: 515.634 ms/op
# Warmup Iteration 4: 496.081 ms/op
# Warmup Iteration 5: 501.787 ms/op
Iteration 1: 495.476 ms/op
Iteration 2: 511.040 ms/op
Iteration 3: 494.175 ms/op
Iteration 4: 493.056 ms/op
Iteration 5: 497.224 ms/op
Result "perf.Java.loop":
497.672 ±(99.9%) 3.576 ms/op [Average]
(min, avg, max) = (492.443, 497.672, 511.165), stdev = 4.773
CI (99.9%): [494.097, 501.248] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Java.stream
# Run progress: 10.00% complete, ETA 01:17:30
# Fork: 1 of 5
# Warmup Iteration 1: 752.945 ms/op
# Warmup Iteration 2: 655.457 ms/op
# Warmup Iteration 3: 656.324 ms/op
# Warmup Iteration 4: 659.984 ms/op
# Warmup Iteration 5: 653.979 ms/op
Iteration 1: 702.193 ms/op
Iteration 2: 658.415 ms/op
Iteration 3: 684.926 ms/op
Iteration 4: 679.830 ms/op
Iteration 5: 682.021 ms/op
# Run progress: 12.00% complete, ETA 01:15:56
# Fork: 2 of 5
# Warmup Iteration 1: 751.638 ms/op
# Warmup Iteration 2: 663.695 ms/op
# Warmup Iteration 3: 663.717 ms/op
# Warmup Iteration 4: 667.699 ms/op
# Warmup Iteration 5: 657.080 ms/op
Iteration 1: 657.204 ms/op
Iteration 2: 669.600 ms/op
Iteration 3: 661.893 ms/op
Iteration 4: 658.053 ms/op
Iteration 5: 677.995 ms/op
# Run progress: 14.00% complete, ETA 01:14:17
# Fork: 3 of 5
# Warmup Iteration 1: 753.697 ms/op
# Warmup Iteration 2: 662.388 ms/op
# Warmup Iteration 3: 665.430 ms/op
# Warmup Iteration 4: 656.421 ms/op
# Warmup Iteration 5: 660.469 ms/op
Iteration 1: 726.873 ms/op
Iteration 2: 675.244 ms/op
Iteration 3: 664.811 ms/op
Iteration 4: 656.496 ms/op
Iteration 5: 655.133 ms/op
# Run progress: 16.00% complete, ETA 01:12:41
# Fork: 4 of 5
# Warmup Iteration 1: 758.258 ms/op
# Warmup Iteration 2: 658.559 ms/op
# Warmup Iteration 3: 653.749 ms/op
# Warmup Iteration 4: 676.368 ms/op
# Warmup Iteration 5: 653.075 ms/op
Iteration 1: 694.018 ms/op
Iteration 2: 663.863 ms/op
Iteration 3: 654.269 ms/op
Iteration 4: 657.201 ms/op
Iteration 5: 682.597 ms/op
# Run progress: 18.00% complete, ETA 01:11:04
# Fork: 5 of 5
# Warmup Iteration 1: 755.350 ms/op
# Warmup Iteration 2: 673.888 ms/op
# Warmup Iteration 3: 655.083 ms/op
# Warmup Iteration 4: 658.737 ms/op
# Warmup Iteration 5: 657.309 ms/op
Iteration 1: 671.508 ms/op
Iteration 2: 673.933 ms/op
Iteration 3: 654.189 ms/op
Iteration 4: 654.793 ms/op
Iteration 5: 658.981 ms/op
Result "perf.Java.stream":
671.042 ±(99.9%) 13.278 ms/op [Average]
(min, avg, max) = (654.189, 671.042, 726.873), stdev = 17.726
CI (99.9%): [657.763, 684.320] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.loop
# Run progress: 20.00% complete, ETA 01:09:23
# Fork: 1 of 5
# Warmup Iteration 1: 501.288 ms/op
# Warmup Iteration 2: 519.426 ms/op
# Warmup Iteration 3: 506.916 ms/op
# Warmup Iteration 4: 494.905 ms/op
# Warmup Iteration 5: 495.065 ms/op
Iteration 1: 505.477 ms/op
Iteration 2: 499.944 ms/op
Iteration 3: 502.763 ms/op
Iteration 4: 494.921 ms/op
Iteration 5: 491.712 ms/op
# Run progress: 22.00% complete, ETA 01:07:35
# Fork: 2 of 5
# Warmup Iteration 1: 519.974 ms/op
# Warmup Iteration 2: 495.087 ms/op
# Warmup Iteration 3: 493.488 ms/op
# Warmup Iteration 4: 503.083 ms/op
# Warmup Iteration 5: 494.544 ms/op
Iteration 1: 498.104 ms/op
Iteration 2: 494.670 ms/op
Iteration 3: 494.472 ms/op
Iteration 4: 496.515 ms/op
Iteration 5: 500.076 ms/op
# Run progress: 24.00% complete, ETA 01:05:49
# Fork: 3 of 5
# Warmup Iteration 1: 502.969 ms/op
# Warmup Iteration 2: 515.297 ms/op
# Warmup Iteration 3: 500.498 ms/op
# Warmup Iteration 4: 504.327 ms/op
# Warmup Iteration 5: 495.414 ms/op
Iteration 1: 492.781 ms/op
Iteration 2: 499.457 ms/op
Iteration 3: 497.536 ms/op
Iteration 4: 506.625 ms/op
Iteration 5: 488.861 ms/op
# Run progress: 26.00% complete, ETA 01:04:02
# Fork: 4 of 5
# Warmup Iteration 1: 499.845 ms/op
# Warmup Iteration 2: 500.112 ms/op
# Warmup Iteration 3: 500.099 ms/op
# Warmup Iteration 4: 502.852 ms/op
# Warmup Iteration 5: 495.422 ms/op
Iteration 1: 520.156 ms/op
Iteration 2: 493.320 ms/op
Iteration 3: 510.862 ms/op
Iteration 4: 497.780 ms/op
Iteration 5: 480.626 ms/op
# Run progress: 28.00% complete, ETA 01:02:15
# Fork: 5 of 5
# Warmup Iteration 1: 507.837 ms/op
# Warmup Iteration 2: 501.298 ms/op
# Warmup Iteration 3: 496.518 ms/op
# Warmup Iteration 4: 499.621 ms/op
# Warmup Iteration 5: 506.226 ms/op
Iteration 1: 492.717 ms/op
Iteration 2: 491.870 ms/op
Iteration 3: 523.607 ms/op
Iteration 4: 496.943 ms/op
Iteration 5: 492.535 ms/op
Result "perf.Kotlin.loop":
498.573 ±(99.9%) 6.962 ms/op [Average]
(min, avg, max) = (480.626, 498.573, 523.607), stdev = 9.294
CI (99.9%): [491.611, 505.535] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.stream
# Run progress: 40.00% complete, ETA 00:51:37
# Fork: 1 of 5
# Warmup Iteration 1: 770.439 ms/op
# Warmup Iteration 2: 654.815 ms/op
# Warmup Iteration 3: 659.349 ms/op
# Warmup Iteration 4: 663.637 ms/op
# Warmup Iteration 5: 657.116 ms/op
Iteration 1: 658.083 ms/op
Iteration 2: 662.591 ms/op
Iteration 3: 674.376 ms/op
Iteration 4: 656.337 ms/op
Iteration 5: 657.640 ms/op
# Run progress: 42.00% complete, ETA 00:49:56
# Fork: 2 of 5
# Warmup Iteration 1: 769.538 ms/op
# Warmup Iteration 2: 654.025 ms/op
# Warmup Iteration 3: 663.950 ms/op
# Warmup Iteration 4: 661.340 ms/op
# Warmup Iteration 5: 721.330 ms/op
Iteration 1: 654.068 ms/op
Iteration 2: 666.005 ms/op
Iteration 3: 658.865 ms/op
Iteration 4: 657.615 ms/op
Iteration 5: 656.568 ms/op
# Run progress: 44.00% complete, ETA 00:48:14
# Fork: 3 of 5
# Warmup Iteration 1: 773.322 ms/op
# Warmup Iteration 2: 655.102 ms/op
# Warmup Iteration 3: 660.459 ms/op
# Warmup Iteration 4: 656.585 ms/op
# Warmup Iteration 5: 715.147 ms/op
Iteration 1: 654.845 ms/op
Iteration 2: 655.245 ms/op
Iteration 3: 662.457 ms/op
Iteration 4: 664.278 ms/op
Iteration 5: 692.333 ms/op
# Run progress: 46.00% complete, ETA 00:46:32
# Fork: 4 of 5
# Warmup Iteration 1: 751.623 ms/op
# Warmup Iteration 2: 655.612 ms/op
# Warmup Iteration 3: 693.453 ms/op
# Warmup Iteration 4: 652.434 ms/op
# Warmup Iteration 5: 664.927 ms/op
Iteration 1: 655.975 ms/op
Iteration 2: 661.740 ms/op
Iteration 3: 661.267 ms/op
Iteration 4: 656.503 ms/op
Iteration 5: 663.981 ms/op
# Run progress: 48.00% complete, ETA 00:44:51
# Fork: 5 of 5
# Warmup Iteration 1: 755.101 ms/op
# Warmup Iteration 2: 659.349 ms/op
# Warmup Iteration 3: 659.407 ms/op
# Warmup Iteration 4: 697.422 ms/op
# Warmup Iteration 5: 653.759 ms/op
Iteration 1: 654.175 ms/op
Iteration 2: 661.572 ms/op
Iteration 3: 653.442 ms/op
Iteration 4: 668.452 ms/op
Iteration 5: 681.106 ms/op
Result "perf.Kotlin.stream":
661.981 ±(99.9%) 6.825 ms/op [Average]
(min, avg, max) = (653.442, 661.981, 692.333), stdev = 9.111
CI (99.9%): [655.156, 668.805] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.streamWrappedInSequence
# Run progress: 50.00% complete, ETA 00:43:08
# Fork: 1 of 5
# Warmup Iteration 1: 4771.596 ms/op
# Warmup Iteration 2: 4469.354 ms/op
# Warmup Iteration 3: 4503.951 ms/op
# Warmup Iteration 4: 4642.394 ms/op
# Warmup Iteration 5: 4570.822 ms/op
Iteration 1: 4734.829 ms/op
Iteration 2: 4448.201 ms/op
Iteration 3: 4450.036 ms/op
Iteration 4: 4649.510 ms/op
Iteration 5: 4360.084 ms/op
# Run progress: 52.00% complete, ETA 00:41:56
# Fork: 2 of 5
# Warmup Iteration 1: 4827.821 ms/op
# Warmup Iteration 2: 4491.528 ms/op
# Warmup Iteration 3: 4429.582 ms/op
# Warmup Iteration 4: 4616.589 ms/op
# Warmup Iteration 5: 4425.327 ms/op
Iteration 1: 4536.462 ms/op
Iteration 2: 4507.816 ms/op
Iteration 3: 4448.700 ms/op
Iteration 4: 4499.069 ms/op
Iteration 5: 4481.440 ms/op
# Run progress: 54.00% complete, ETA 00:40:38
# Fork: 3 of 5
# Warmup Iteration 1: 5100.392 ms/op
# Warmup Iteration 2: 4510.386 ms/op
# Warmup Iteration 3: 4520.391 ms/op
# Warmup Iteration 4: 4455.398 ms/op
# Warmup Iteration 5: 4438.041 ms/op
Iteration 1: 4463.642 ms/op
Iteration 2: 4513.981 ms/op
Iteration 3: 4451.207 ms/op
Iteration 4: 4592.493 ms/op
Iteration 5: 4558.072 ms/op
# Run progress: 56.00% complete, ETA 00:39:12
# Fork: 4 of 5
# Warmup Iteration 1: 4625.825 ms/op
# Warmup Iteration 2: 4523.034 ms/op
# Warmup Iteration 3: 4642.197 ms/op
# Warmup Iteration 4: 4420.747 ms/op
# Warmup Iteration 5: 4655.348 ms/op
Iteration 1: 4655.329 ms/op
Iteration 2: 4426.920 ms/op
Iteration 3: 4483.991 ms/op
Iteration 4: 4499.432 ms/op
Iteration 5: 4537.659 ms/op
# Run progress: 58.00% complete, ETA 00:37:47
# Fork: 5 of 5
# Warmup Iteration 1: 5112.802 ms/op
# Warmup Iteration 2: 4446.181 ms/op
# Warmup Iteration 3: 4580.273 ms/op
# Warmup Iteration 4: 4410.598 ms/op
# Warmup Iteration 5: 4403.912 ms/op
Iteration 1: 4502.871 ms/op
Iteration 2: 4654.358 ms/op
Iteration 3: 4422.066 ms/op
Iteration 4: 4356.905 ms/op
Iteration 5: 4456.772 ms/op
Result "perf.Kotlin.streamWrappedInSequence":
4507.674 ±(99.9%) 69.124 ms/op [Average]
(min, avg, max) = (4356.905, 4507.674, 4734.829), stdev = 92.279
CI (99.9%): [4438.549, 4576.798] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.Kotlin.withGenerator
# Run progress: 60.00% complete, ETA 00:36:14
# Fork: 1 of 5
# Warmup Iteration 1: 12972.312 ms/op
# Warmup Iteration 2: 11150.766 ms/op
# Warmup Iteration 3: 9243.075 ms/op
# Warmup Iteration 4: 9264.701 ms/op
# Warmup Iteration 5: 11170.115 ms/op
Iteration 1: 9341.833 ms/op
Iteration 2: 9283.496 ms/op
Iteration 3: 9174.673 ms/op
Iteration 4: 9202.388 ms/op
Iteration 5: 9320.724 ms/op
# Run progress: 62.00% complete, ETA 00:35:07
# Fork: 2 of 5
# Warmup Iteration 1: 11221.832 ms/op
# Warmup Iteration 2: 14477.742 ms/op
# Warmup Iteration 3: 9534.827 ms/op
# Warmup Iteration 4: 9291.195 ms/op
# Warmup Iteration 5: 9361.180 ms/op
Iteration 1: 13425.934 ms/op
Iteration 2: 9402.840 ms/op
Iteration 3: 9205.030 ms/op
Iteration 4: 9145.462 ms/op
Iteration 5: 9412.641 ms/op
# Run progress: 64.00% complete, ETA 00:33:49
# Fork: 3 of 5
# Warmup Iteration 1: 11481.680 ms/op
# Warmup Iteration 2: 11080.843 ms/op
# Warmup Iteration 3: 9253.032 ms/op
# Warmup Iteration 4: 9399.855 ms/op
# Warmup Iteration 5: 8943.780 ms/op
Iteration 1: 9482.955 ms/op
Iteration 2: 9124.821 ms/op
Iteration 3: 9198.032 ms/op
Iteration 4: 9194.677 ms/op
Iteration 5: 9252.317 ms/op
# Run progress: 66.00% complete, ETA 00:32:26
# Fork: 4 of 5
# Warmup Iteration 1: 11560.755 ms/op
# Warmup Iteration 2: 10979.804 ms/op
# Warmup Iteration 3: 8778.534 ms/op
# Warmup Iteration 4: 9271.147 ms/op
# Warmup Iteration 5: 8962.573 ms/op
Iteration 1: 8849.864 ms/op
Iteration 2: 8943.724 ms/op
Iteration 3: 8956.114 ms/op
Iteration 4: 8851.836 ms/op
Iteration 5: 8918.313 ms/op
# Run progress: 68.00% complete, ETA 00:30:56
# Fork: 5 of 5
# Warmup Iteration 1: 11657.552 ms/op
# Warmup Iteration 2: 14300.899 ms/op
# Warmup Iteration 3: 9345.045 ms/op
# Warmup Iteration 4: 9167.124 ms/op
# Warmup Iteration 5: 9481.174 ms/op
Iteration 1: 9249.084 ms/op
Iteration 2: 9231.042 ms/op
Iteration 3: 9360.817 ms/op
Iteration 4: 9328.045 ms/op
Iteration 5: 9094.625 ms/op
Result "perf.Kotlin.withGenerator":
9358.051 ±(99.9%) 647.987 ms/op [Average]
(min, avg, max) = (8849.864, 9358.051, 13425.934), stdev = 865.044
CI (99.9%): [8710.065, 10006.038] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.parallelStream
# Run progress: 70.00% complete, ETA 00:29:25
# Fork: 1 of 5
# Warmup Iteration 1: 45622.300 ms/op
# Warmup Iteration 2: 51268.916 ms/op
# Warmup Iteration 3: 54557.341 ms/op
# Warmup Iteration 4: 53486.933 ms/op
# Warmup Iteration 5: 49141.534 ms/op
Iteration 1: 48819.449 ms/op
Iteration 2: 49136.332 ms/op
Iteration 3: 50845.214 ms/op
Iteration 4: 49430.190 ms/op
Iteration 5: 48870.124 ms/op
# Run progress: 72.00% complete, ETA 00:29:57
# Fork: 2 of 5
# Warmup Iteration 1: 50170.272 ms/op
# Warmup Iteration 2: 53994.209 ms/op
# Warmup Iteration 3: 53407.886 ms/op
# Warmup Iteration 4: 51364.240 ms/op
# Warmup Iteration 5: 51227.612 ms/op
Iteration 1: 51931.093 ms/op
Iteration 2: 50769.543 ms/op
Iteration 3: 56751.344 ms/op
Iteration 4: 52092.542 ms/op
Iteration 5: 53091.324 ms/op
# Run progress: 74.00% complete, ETA 00:30:08
# Fork: 3 of 5
# Warmup Iteration 1: 49080.774 ms/op
# Warmup Iteration 2: 56156.551 ms/op
# Warmup Iteration 3: 52052.861 ms/op
# Warmup Iteration 4: 55415.614 ms/op
# Warmup Iteration 5: 51782.162 ms/op
Iteration 1: 51509.481 ms/op
Iteration 2: 55253.665 ms/op
Iteration 3: 50771.941 ms/op
Iteration 4: 54707.276 ms/op
Iteration 5: 51359.288 ms/op
# Run progress: 76.00% complete, ETA 00:29:52
# Fork: 4 of 5
# Warmup Iteration 1: 50272.019 ms/op
# Warmup Iteration 2: 57595.569 ms/op
# Warmup Iteration 3: 59850.768 ms/op
# Warmup Iteration 4: 54712.763 ms/op
# Warmup Iteration 5: 53759.817 ms/op
Iteration 1: 55310.711 ms/op
Iteration 2: 52339.983 ms/op
Iteration 3: 53796.593 ms/op
Iteration 4: 51265.637 ms/op
Iteration 5: 51163.343 ms/op
# Run progress: 78.00% complete, ETA 00:29:13
# Fork: 5 of 5
# Warmup Iteration 1: 46287.662 ms/op
# Warmup Iteration 2: 56583.661 ms/op
# Warmup Iteration 3: 54958.161 ms/op
# Warmup Iteration 4: 52485.864 ms/op
# Warmup Iteration 5: 51463.926 ms/op
Iteration 1: 59061.271 ms/op
Iteration 2: 62958.959 ms/op
Iteration 3: 54422.187 ms/op
Iteration 4: 51840.252 ms/op
Iteration 5: 52454.721 ms/op
Result "perf.NonPrimitive.parallelStream":
52798.098 ±(99.9%) 2451.723 ms/op [Average]
(min, avg, max) = (48819.449, 52798.098, 62958.959), stdev = 3272.980
CI (99.9%): [50346.375, 55249.822] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.sequence
# Run progress: 80.00% complete, ETA 00:28:09
# Fork: 1 of 5
# Warmup Iteration 1: 31539.028 ms/op
# Warmup Iteration 2: 32167.361 ms/op
# Warmup Iteration 3: 31438.191 ms/op
# Warmup Iteration 4: 31771.676 ms/op
# Warmup Iteration 5: 31476.264 ms/op
Iteration 1: 32394.913 ms/op
Iteration 2: 31534.739 ms/op
Iteration 3: 32371.211 ms/op
Iteration 4: 31587.742 ms/op
Iteration 5: 32116.086 ms/op
# Run progress: 82.00% complete, ETA 00:25:53
# Fork: 2 of 5
# Warmup Iteration 1: 32147.157 ms/op
# Warmup Iteration 2: 31512.206 ms/op
# Warmup Iteration 3: 31765.894 ms/op
# Warmup Iteration 4: 31205.766 ms/op
# Warmup Iteration 5: 32011.824 ms/op
Iteration 1: 31948.070 ms/op
Iteration 2: 35102.995 ms/op
Iteration 3: 31417.171 ms/op
Iteration 4: 31561.729 ms/op
Iteration 5: 31358.093 ms/op
# Run progress: 84.00% complete, ETA 00:23:29
# Fork: 3 of 5
# Warmup Iteration 1: 31004.566 ms/op
# Warmup Iteration 2: 31665.419 ms/op
# Warmup Iteration 3: 31381.758 ms/op
# Warmup Iteration 4: 31739.301 ms/op
# Warmup Iteration 5: 33043.687 ms/op
Iteration 1: 31251.636 ms/op
Iteration 2: 34870.238 ms/op
Iteration 3: 31329.340 ms/op
Iteration 4: 31706.639 ms/op
Iteration 5: 31342.416 ms/op
# Run progress: 86.00% complete, ETA 00:20:56
# Fork: 4 of 5
# Warmup Iteration 1: 31439.350 ms/op
# Warmup Iteration 2: 31796.581 ms/op
# Warmup Iteration 3: 30981.395 ms/op
# Warmup Iteration 4: 31014.350 ms/op
# Warmup Iteration 5: 32090.784 ms/op
Iteration 1: 35625.972 ms/op
Iteration 2: 31931.136 ms/op
Iteration 3: 31184.763 ms/op
Iteration 4: 31694.722 ms/op
Iteration 5: 31880.194 ms/op
# Run progress: 88.00% complete, ETA 00:18:16
# Fork: 5 of 5
# Warmup Iteration 1: 32086.933 ms/op
# Warmup Iteration 2: 31691.944 ms/op
# Warmup Iteration 3: 31720.550 ms/op
# Warmup Iteration 4: 31687.710 ms/op
# Warmup Iteration 5: 31665.377 ms/op
Iteration 1: 31649.628 ms/op
Iteration 2: 31965.411 ms/op
Iteration 3: 31552.680 ms/op
Iteration 4: 31191.134 ms/op
Iteration 5: 31948.312 ms/op
Result "perf.NonPrimitive.sequence":
32100.679 ±(99.9%) 913.441 ms/op [Average]
(min, avg, max) = (31184.763, 32100.679, 35625.972), stdev = 1219.417
CI (99.9%): [31187.238, 33014.120] (assumes normal distribution)
# JMH version: 1.23
# VM version: JDK 15-ea, OpenJDK 64-Bit Server VM, 15-ea+17-717
# VM invoker: /usr/lib/jvm/java-15-openjdk/bin/java
# VM options: <none>
# Warmup: 5 iterations, 10 s each
# Measurement: 5 iterations, 10 s each
# Timeout: 10 min per iteration
# Threads: 1 thread, will synchronize iterations
# Benchmark mode: Average time, time/op
# Benchmark: perf.NonPrimitive.stream
# Run progress: 90.00% complete, ETA 00:15:28
# Fork: 1 of 5
# Warmup Iteration 1: 43230.332 ms/op
# Warmup Iteration 2: 50557.338 ms/op
# Warmup Iteration 3: 50822.458 ms/op
# Warmup Iteration 4: 50159.518 ms/op
# Warmup Iteration 5: 50098.131 ms/op
Iteration 1: 50816.292 ms/op
Iteration 2: 50245.734 ms/op
Iteration 3: 50341.716 ms/op
Iteration 4: 50361.851 ms/op
Iteration 5: 50362.410 ms/op
# Run progress: 92.00% complete, ETA 00:12:49
# Fork: 2 of 5
# Warmup Iteration 1: 53187.949 ms/op
# Warmup Iteration 2: 51120.670 ms/op
# Warmup Iteration 3: 51116.071 ms/op
# Warmup Iteration 4: 51517.877 ms/op
# Warmup Iteration 5: 51755.674 ms/op
Iteration 1: 51481.433 ms/op
Iteration 2: 51822.715 ms/op
Iteration 3: 51509.446 ms/op
Iteration 4: 51537.729 ms/op
Iteration 5: 53196.225 ms/op
# Run progress: 94.00% complete, ETA 00:09:58
# Fork: 3 of 5
# Warmup Iteration 1: 43924.846 ms/op
# Warmup Iteration 2: 50782.107 ms/op
# Warmup Iteration 3: 50232.693 ms/op
# Warmup Iteration 4: 49835.131 ms/op
# Warmup Iteration 5: 50901.686 ms/op
Iteration 1: 50166.901 ms/op
Iteration 2: 50617.046 ms/op
Iteration 3: 50144.998 ms/op
Iteration 4: 50044.519 ms/op
Iteration 5: 50792.107 ms/op
# Run progress: 96.00% complete, ETA 00:06:51
# Fork: 4 of 5
# Warmup Iteration 1: 44810.711 ms/op
# Warmup Iteration 2: 51014.134 ms/op
# Warmup Iteration 3: 50839.319 ms/op
# Warmup Iteration 4: 51794.822 ms/op
# Warmup Iteration 5: 51088.069 ms/op
Iteration 1: 51650.679 ms/op
Iteration 2: 51178.894 ms/op
Iteration 3: 51247.988 ms/op
Iteration 4: 50767.387 ms/op
Iteration 5: 50682.972 ms/op
# Run progress: 98.00% complete, ETA 00:03:31
# Fork: 5 of 5
# Warmup Iteration 1: 42961.998 ms/op
# Warmup Iteration 2: 50554.667 ms/op
# Warmup Iteration 3: 49991.236 ms/op
# Warmup Iteration 4: 49157.238 ms/op
# Warmup Iteration 5: 50339.484 ms/op
Iteration 1: 50025.853 ms/op
Iteration 2: 50515.643 ms/op
Iteration 3: 50206.781 ms/op
Iteration 4: 49584.429 ms/op
Iteration 5: 52719.464 ms/op
Result "perf.NonPrimitive.stream":
50880.849 ±(99.9%) 644.167 ms/op [Average]
(min, avg, max) = (49584.429, 50880.849, 53196.225), stdev = 859.945
CI (99.9%): [50236.681, 51525.016] (assumes normal distribution)
# Run complete. Total time: 03:01:10
REMEMBER: The numbers below are just data. To gain reusable insights, you need to follow up on
why the numbers are the way they are. Use profilers (see -prof, -lprof), design factorial
experiments, perform baseline and negative tests that provide experimental control, make sure
the benchmarking environment is safe on JVM/OS/HW level, ask for reviews from the domain experts.
Do not assume the numbers tell you what you want them to tell.
Benchmark Mode Cnt Score Error Units
Java.loop avgt 25 497.672 ± 3.576 ms/op
Java.stream avgt 25 671.042 ± 13.278 ms/op
Kotlin.loop avgt 25 498.573 ± 6.962 ms/op
Kotlin.stream avgt 25 661.981 ± 6.825 ms/op
Kotlin.streamWrappedInSequence avgt 25 4507.674 ± 69.124 ms/op
Kotlin.withGenerator avgt 25 9358.051 ± 647.987 ms/op
NonPrimitive.parallelStream avgt 25 52798.098 ± 2451.723 ms/op
NonPrimitive.sequence avgt 25 32100.679 ± 913.441 ms/op
NonPrimitive.stream avgt 25 50880.849 ± 644.167 ms/op

@ -1,4 +1,4 @@
# Stream, Sequence, Iterator – a story of laziness and sad benchmarking noises
# Stream, Sequence, Iterator – a story of laziness and sad JVM benchmarking noises
Many programming languages have started to include more functional features in their standard libraries.
One of those features is lazy collections, for lack of a better term,
which seem to have a different name in each language
@ -12,39 +12,359 @@ For almost every language out there that offers lazy iterators,
I was personally interested in this because, being a Java/Kotlin developer,
I use Java’s Streams and Kotlin’s Sequences almost every day
with relatively little regard for performance.
with relatively little regard for potential performance implications.
They are intuitive to write and are easy to reason about,
which is usually much more important than the results of a thousand microbenchmarks,
so please don’t stop using your favorite language feature because it’s 2.8% slower than the primitive alternative.
so please don’t stop using your favorite language feature because it’s 2.8% slower than the alternative.
Most code is already bad enough as is without desperate optimization attempts.
Still, I wanted to know how they compare to imperative code.
There are some resources on this for Java 8’s Stream API,
but Kotlin’s Sequences seem to just be accepted as
more convenient Streams[^convenience].
Rust is here as a baseline for comparisons
because it is generally regarded as having very optimized iterators.
[^convenience]: If you’ve ever used them, you’ll know what I mean.
Java’s Streams are built in a way that allows for easy parallelism,
but brings its own problems and limitations for the usage.
We’ll see some of that in the code examples later.
## What *is* an iterator?
You can think of an iterator as a pipeline of data.
It’s not a list, so it doesn’t support indexing,
because it doesn’t actually hold any data.
It just contains information on how to get or make that data.
You can make it produce data and use that
(which is often called ‘consuming’ the iterator
because if you read data from the pipeline, it’s usually gone),
or you can add a new step to the pipeline and hand the result to someone else,
who can then consume it or add even more operations.
You also don’t know when (or if at all) an iterator will end.
Someone could sit at the other end and constantly put new data into your pipeline.
You can think of an iterator as a pipeline.
It lets you write code as a sequence of instructions to be applied to all elements of a container.
An important aspect to note is: adding an operation to the pipeline does nothing
Let’s use a simple example to demonstrate this.
We want to take all numbers from 1 to 100,000,
multiply each of them by 2,
and then add all of them.[^sum]
[^sum]: You could also just compute the sum and take that \* 2, but we specifically want that intermediate step for the example.
First, the imperative solution:
```kotlin
var sum = 0
for (i in 1..100_000) {
sum += i * 2
}
return sum
```
and now the functional version using a Sequence (Kotlin’s name for Streams/iterators):
```kotlin
return (1..100_000).asSequence()
.map { it * 2 }
.sum()
```
An iterator not a list, and it doesn’t support indexing,[^index]
because it doesn’t actually contain any data.
It just knows how to get or compute it for you,
but you don’t know how it does that.
You don’t even always know when (or if at all) an iterator will end
(in this case, we do, because we create the Sequence from the range `1..100_000`,
meaning it will produce the numbers from 1 to 100,00 before it ends).
You can tell an iterator to produce or emit data if you want to use it
(which is often called ‘consuming’
because if you read data from the pipeline,
it’s usually gone),
or you can add a new step to it and hand the new pipeline to someone else,
who can then consume it or add even more steps.
[^index]: Or any other operation like it.
No `iterator[0]`, no `iterator.get(0)` or whatever your favorite language uses.
An operation like `iterator.last()` might exist,
but it will consume the entire iterator instead of just accessing the last element.
An important aspect to note is:
adding an operation to the pipeline does nothing
until someone actually starts reading from it,
and even then, only the elements that are consumed are computed.[^inf]
and even then, only the elements that are consumed are computed.[^inf]
This makes it possible to operate on huge data sets while keeping memory usage low,
because only the currently active element has to be held in memory.
[^inf]: This is what makes infinite iterators possible to begin with.
[^inf]: This is what makes infinite iterators possible.
They can be very useful and are used a lot in functional languages,
but they’re not today’s topic.
## Cold, hard numbers
We’ll use that small example from the last section as our first example:
take a range of numbers, double each number, and compute the sum –
except this time, we’ll do the numbers from 1 to 1 billion.
Since everything we’re doing is lazy, memory usage shouldn’t be an issue.
I will use different implementations to solve them and benchmark all of them.
Here are the different approaches I came up with:
- a simple for loop in Java
- Java’s LongStream
- a for each loop with a range in Kotlin
- Java’s LongStream called from Kotlin[^ktjava]
- Java’s Stream wrapped in a Kotlin Sequence
- a Kotlin range wrapped in a Sequence
- Kotlin’s Sequence with a generator to create the range
[^ktjava]: Mainly to make sure there is no performance difference between the two.
The benchmarks were executed on an Intel Xeon E3-1271 v3 with 32 GB of RAM,
running Arch Linux with kernel 5.4.20-1-lts,
using the (at the time of writing) latest OpenJDK preview build (`15-ea+17-717`),
Kotlin 1.4-M1, and [jmh](https://openjdk.java.net/projects/code-tools/jmh/) version 1.23.
The bytecode target was set to Java 15 for the Java code and Java 13 for Kotlin
(newer versions are currently unsupported).
Source code for the Java tests:
```java
public long stream() {
return LongStream.range(1, upper)
.map(l -> l * 2)
.sum();
}
public long loop() {
long sum = 0;
for (long i = 0; i < upper; i++) {
sum += i * 2;
}
return sum;
}
```
and for Kotlin:
```kotlin
fun stream() =
LongStream.range(1, upper)
.map { it * 2 }
.sum()
fun loop(): Long {
var sum = 0L
for (l in 1L until upper) {
sum += l * 2
}
return sum
}
fun streamWrappedInSequence() =
LongStream.range(1L, upper)
.asSequence()
.map { it * 2 }
.sum()
fun sequence() =
(1 until upper).asSequence()
.map { it * 2 }
.sum()
fun withGenerator() =
generateSequence(0L, { it + 1L })
.take(upper.toInt())
.map { it * 2 }
.sum()
```
with `const val upper = 1_000_000_000L`.[^`1 until upper` is used in these examples because unlike `lower..upper`, `until` is end-inclusive like Java’s LongStream.range().]
Without wasting any more of your time, here are the results:
```plaintext
Benchmark Mode Cnt Score Error Units
Java.loop avgt 25 446.055 ± 0.677 ms/op
Java.stream avgt 25 601.424 ± 12.606 ms/op
Kotlin.loop avgt 25 446.600 ± 1.164 ms/op
Kotlin.sequence avgt 25 2732.604 ± 6.644 ms/op
Kotlin.stream avgt 25 593.353 ± 1.408 ms/op
Kotlin.streamWrappedInSequence avgt 25 3829.209 ± 33.569 ms/op
Kotlin.withGenerator avgt 25 8374.149 ± 880.647 ms/op
```
(full JMH output)[https://ruru.moe/pSK13p8].
Unsurprisingly, using Streams from Java and Kotlin is almost identical in terms of performance.
The same is true for imperative loops,
meaning Kotlin ranges introduce no overhead compared to incrementing for loops.
More surprisingly, using Sequences is an order of magnitude slower.
That was not at all according to my expectations, so I investigated.
As it turns out, Java’s `LongStream` exists because Stream<Long> is *much* slower.
The JVM has to use `Long` rather than `long` when the type is used for generics,
which involves an additional boxing step and the allocation for the `Long` object.[^primitives]
Still, we now know that Streams have about 25% overhead compared to the simple loop for this example,
that generating sequences is a comparatively slow process,
and that wrapping Streams comes at a considerable cost (compared to a sequence created from a range).
[^primitives]: The JVM has a few primitive types, such as `int`, `char`, or array types.
They are different from any other type because they cannot be `null`.
Every regular type on the JVM extends `java.lang.Object` and is just a reference that is being passed around.
The primitives are values, not references, so there’s a lot less overhead involved.
Unfortunately, primitives can’t be used as generic types,
so a list of longs will always convert the `long` to `Long` before adding it.
That last point seemed odd, so I attached a profiler to see where the CPU time is lost.
![Flamegraph of `streamWrappedInSequence()`](https://i.kageru.moe/knT2Eg.png)
We can see that the `LongStream` can produce a `PrimitiveIterator.OfLong` that is used as a source for the Sequence.
The operation of boxing a primitive `long` into an object `Long`
(that’s the `Long.valueOf()` step) takes almost as long as advancing the underlying iterator itself.
7.7% of the CPU time is spent in `Sequence.hasNext()`.
The exact breakdown of that looks as follows:
![Checking if a Sequence has more elements](https://i.kageru.moe/k4NHhR.png)
The Sequence introduces very little overhead here, as it just delegates to `hasNext()` of the underlying iterator.
Worth noting is that the iterator calls `accept` as part of `hasNext()`,
which will already advance the underlying iterator.
The value returned by that will be stored temporarily until `nextLong()` is called.
```java
public boolean tryAdvance(LongConsumer consumer) {
final long i = from;
if (i < upTo) {
from++;
consumer.accept(i);
return true;
}
// more stuff down here
}
```
where `consumer.accept()` is
```java
public void accept(T t) {
valueReady = true;
nextElement = t;
}
```
Knowing this, I have to wonder why `nextLong()` takes as long as it does.
Looking at [the implementation](https://github.com/openjdk/jdk/blob/6bab0f539fba8fb441697846347597b4a0ade428/src/java.base/share/classes/java/util/Spliterators.java#L756),
I don’t understand where all that time is going.
`hasNext()` should always be called before `next()`,
so `next()` just has to return a precomputed value.
Nevertheless, we can now explain the performance difference with the additional boxing step.
Primitives good; everything else bad.
With that in mind, I wrote a second test that avoids the unboxing issue to compare Streams and Sequences.
The next snippet uses a simple wrapper class that guarantees that we have no primitives
to execute a few operations on a Stream/Sequence.
I’ll use this opportunity to also compare parallel and sequential streams.
The steps are simple:
Take a long -> create a LongWrapper from it, double the contained value (which creates a new LongWrapper), extract the value, compute the sum.
That may sound overcomplicated,
but it’s sadly close to the reality of enterprise code.
Wrapper types are everywhere.
```kotlin
inner class LongWrapper(val value: Long) {
fun double() = LongWrapper(value * 2)
}
fun sequence(): Long =
(1 until upper).asSequence()
.map(::LongWrapper)
.map(LongWrapper::double)
.map(LongWrapper::value)
.sum()
fun stream(): Optional<Long> =
StreamSupport.stream((1 until upper).spliterator(), false)
.map(::LongWrapper)
.map(LongWrapper::double)
.map(LongWrapper::value)
.reduce(Long::plus)
fun parallelStream(): Optional<Long> =
StreamSupport.stream((1 until upper).spliterator(), true)
.map(::LongWrapper)
.map(LongWrapper::double)
.map(LongWrapper::value)
.reduce(Long::plus)
fun loop(): Long {
var sum = 0L
for (l in 1 until upper) {
val wrapper = LongWrapper(l)
val doubled = wrapper.double()
sum += doubled.value
}
return sum
}
```
The results here paint a different picture:
```
NonPrimitive.loop avgt 25 445.992 ± 0.642 ms/op
NonPrimitive.sequence avgt 25 27257.399 ± 342.686 ms/op
NonPrimitive.stream avgt 25 44673.318 ± 1325.832 ms/op
NonPrimitive.parallelStream avgt 25 33856.919 ± 249.911 ms/op
```
Full results are in the (JMH log from earlier)[https://ruru.moe/pSK13p8].
The overhead of Java streams is much higher than that of Kotlin Sequences,
and even a parallel Stream is slower than using a Sequence.
even though Sequences only use a single thread,
but both are miles behind the simple for loop.
My first assumption was that the compiler optimized away the wrapper type and just added the longs,
but looking at [the bytecode](https://p.kageru.moe/AUJKiG),
the constructor invocation and the `double()` method calls are still there.
It’s hard to know what the JIT does at runtime,
but the numbers certainly suggest that the wrapper is simply optimized away.
The profiler report wasn’t helpful either,
which further leads me to believe that the JIT just deletes the method and inlines the calculations.
This tells us that not only do Streams/Sequences have a very measurable overhead,
but they severely limit the optimizer’s
(be it compile-time or JIT)
ability to understand the code,
which can lead to significant slowdowns in code that can be optimized.
Obviously, code that doesn’t rely on the optimizer as much won’t be affected to the same degree.
## Conclusion
Overall, I think that Kotlin’s Sequences are a good addition to the language.
They fall behind Streams when working with primitives
because the Java standard library has subtypes for many generic constructs to more efficiently handle primitive types,
but in most real-world JVM applications (that being enterprise-level bloatware),
primitives are the exception rather than the rule.
Still, Kotlin has some types that optimize for this, such as `LongIterator`,
but without a `LongSequence` to go with it,
the boxing will still happen eventually,
and all the performance gains are void.
I hope that we can get a few more types like it in the future,
which will be especially useful once Kotlin/Native reaches maturity
and starts being used for small/embedded hardware.
Apart from the performance, Sequences are also a lot easier to understand and even extend than Streams.
Implementing your own Sequence requires barely more than an implementation of the underlying iterator,
as can be seen in [CoalescingSequence](https://git.kageru.moe/kageru/Sekwences/src/branch/master/src/main/kotlin/moe/kageru/sekwences/CoalescingSequence.kt)
which I implemented last year to get a feeling for how all of this works.
Streams on the other hand are a lot more complex. They actually extend `Consumer<T>`,
so a `Stream<T>` is just a `void consume(T input)` that can be called repeatedly.
That makes it a lot harder to grasp where data is coming from and how it is requested, at least to me.
Simplicity is often underrated in software, but I consider it a huge plus for Sequences.
I will continue to use them liberally,
unless I find myself in a situation where I need to process a huge number of primitives.
And even then, I now know that Java’s Streams are a good choice.
25% might sound like a lot,
but it’s more than worth it if it means leaving code that is much easier to understand and modify for the next person.
Unless you’re actually in a very performance-critical part of your application,
but if you ever find yourself in that situation,
you should switch to a different language.
On that note: I was originally going to include Rust’s iterators here for comparison,
but rustc optimized away all of my benchmarks with [constant time solutions](https://godbolt.org/z/iJaWVP).
It’s a fascinating topic,
and I might write a separate blog post
where I dissect some of the assembly that rustc/LLVM produced,
but I need to properly understand it myself first.