Dev Builds » 20200127-1016

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 15. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo
ncm-dbt-01 06:27:20 581865 4000 50 2572 1378 -257.88 ± 6.54 597 1328 75 0 0 -687.51 ± 39.96
ncm-dbt-02 06:27:27 586544 4000 67 2569 1364 -255.01 ± 6.55 587 1329 83 1 0 -665.33 ± 37.93
ncm-dbt-03 06:27:45 587564 4000 50 2601 1349 -262.1 ± 6.71 623 1307 68 2 0 -694.74 ± 42.02
ncm-dbt-04 06:28:09 566891 4000 42 2554 1404 -256.44 ± 6.42 581 1351 67 1 0 -702.26 ± 42.37
ncm-dbt-05 06:29:10 580263 4000 57 2548 1395 -253.45 ± 6.39 569 1354 76 1 0 -680.57 ± 39.69
20000 266 12844 6890 -256.96 ± 2.92 2957 6669 369 5 0 -685.63 ± 17.78

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
445131 ncm-dbt-05 577602 500 5 315 180 -251.89 ± 16.53 65 180 5 0 0 -798.18 ± 231.48
445130 ncm-dbt-04 567601 500 8 314 178 -247.41 ± 17.1 65 176 9 0 0 -694.7 ± 134.18
445129 ncm-dbt-01 578712 500 8 328 164 -263.41 ± 20.22 83 154 13 0 0 -629.41 ± 104.8
445128 ncm-dbt-03 587664 500 6 328 166 -265.77 ± 19.13 80 162 8 0 0 -715.51 ± 146.53
445127 ncm-dbt-02 585253 500 11 316 173 -246.3 ± 18.81 70 166 13 1 0 -603.86 ± 103.96
445126 ncm-dbt-05 577807 500 7 325 168 -261.07 ± 19.11 78 162 10 0 0 -676.04 ± 124.55
445125 ncm-dbt-02 585632 500 8 311 181 -244.09 ± 16.95 63 177 10 0 0 -676.04 ± 124.55
445124 ncm-dbt-04 566138 500 7 322 171 -257.59 ± 18.4 74 167 9 0 0 -694.7 ± 134.18
445123 ncm-dbt-01 581444 500 5 324 171 -262.24 ± 18.98 78 163 9 0 0 -694.7 ± 134.18
445122 ncm-dbt-03 588387 500 9 327 164 -261.07 ± 19.38 79 160 11 0 0 -659.13 ± 116.77
445121 ncm-dbt-05 582360 500 10 310 180 -240.82 ± 17.32 63 174 13 0 0 -629.41 ± 104.8
445120 ncm-dbt-02 587877 500 6 327 167 -264.59 ± 19.27 80 161 9 0 0 -694.7 ± 134.18
445119 ncm-dbt-04 569589 500 5 313 182 -249.64 ± 18.21 70 168 12 0 0 -643.67 ± 110.29
445118 ncm-dbt-03 586350 500 6 322 172 -258.74 ± 19.09 77 162 11 0 0 -659.13 ± 116.77
445117 ncm-dbt-01 584369 500 8 316 176 -249.64 ± 18.21 70 168 12 0 0 -643.67 ± 110.29
445116 ncm-dbt-05 580365 500 5 314 181 -250.76 ± 16.68 65 179 6 0 0 -766.17 ± 188.02
445115 ncm-dbt-02 587877 500 9 327 164 -261.07 ± 19.11 78 162 10 0 0 -676.04 ± 124.55
445114 ncm-dbt-04 567007 500 2 333 165 -276.68 ± 18.94 84 163 3 0 0 -887.58 ± 222.18
445113 ncm-dbt-03 588132 500 4 330 166 -270.57 ± 19.13 82 162 6 0 0 -766.17 ± 188.02
445112 ncm-dbt-01 580406 500 8 311 181 -244.09 ± 17.22 64 175 11 0 0 -659.13 ± 116.77
445111 ncm-dbt-05 582068 500 7 330 163 -266.97 ± 20.66 85 154 10 1 0 -643.7 ± 122.54
445110 ncm-dbt-04 564801 500 6 317 177 -253.02 ± 18.9 73 166 10 1 0 -643.7 ± 122.54
445109 ncm-dbt-02 585717 500 4 322 174 -261.07 ± 17.97 74 170 6 0 0 -766.17 ± 188.02
445108 ncm-dbt-03 588600 500 8 329 163 -264.59 ± 19.55 81 159 10 0 0 -676.04 ± 124.55
445107 ncm-dbt-01 581444 500 3 314 183 -253.02 ± 16.37 65 181 4 0 0 -837.28 ± 240.94
445106 ncm-dbt-01 581777 500 5 336 159 -276.68 ± 20.16 88 155 7 0 0 -739.05 ± 163.27
445105 ncm-dbt-05 582861 500 9 318 173 -250.76 ± 18.35 71 167 12 0 0 -643.67 ± 110.29
445104 ncm-dbt-02 583908 500 11 325 164 -256.44 ± 20.12 80 154 16 0 0 -592.27 ± 92.24
445103 ncm-dbt-04 565863 500 6 327 167 -264.59 ± 19.27 80 161 9 0 0 -694.7 ± 134.18
445102 ncm-dbt-03 586350 500 5 320 175 -257.59 ± 18.68 75 165 10 0 0 -676.04 ± 124.55
445101 ncm-dbt-05 577848 500 7 325 168 -261.07 ± 18.83 77 164 9 0 0 -694.7 ± 134.18
445100 ncm-dbt-01 583279 500 8 327 165 -262.24 ± 19.26 79 161 10 0 0 -676.04 ± 124.55
445099 ncm-dbt-03 588004 500 7 306 187 -239.74 ± 17.17 60 181 7 2 0 -659.17 ± 146.32
445098 ncm-dbt-04 567007 500 6 315 179 -250.76 ± 17.53 68 173 9 0 0 -694.7 ± 134.18
445097 ncm-dbt-02 587707 500 5 327 168 -265.77 ± 18.55 78 166 6 0 0 -766.17 ± 188.02
445096 ncm-dbt-01 583489 500 5 316 179 -253.02 ± 17.82 70 171 9 0 0 -694.7 ± 134.18
445095 ncm-dbt-03 587028 500 5 339 156 -280.42 ± 20.0 89 156 5 0 0 -798.18 ± 231.48
445094 ncm-dbt-05 581194 500 7 311 182 -245.19 ± 17.36 65 174 11 0 0 -659.13 ± 116.77
445093 ncm-dbt-02 588387 500 13 314 173 -241.91 ± 17.46 64 173 13 0 0 -629.41 ± 104.8
445092 ncm-dbt-04 567126 500 2 313 185 -253.02 ± 16.96 67 177 6 0 0 -766.17 ± 188.02

Commit

Commit ID 39437f4e55aaa26ef9f0d5a1c762e560e9ffde32
Author Sami Kiminki
Date 2020-01-27 10:16:10 UTC
Advise the kernel to use huge pages (Linux) Align the TT allocation by 2M to make it huge page friendly and advise the kernel to use huge pages. Benchmarks on my i7-8700K (6C/12T) box: (3 runs per bench per config) vanilla (nps) hugepages (nps) avg ================================================================================== bench | 3012490 3024364 3036331 3071052 3067544 3071052 +1.5% bench 16 12 20 | 19237932 19050166 19085315 19266346 19207025 19548758 +1.1% bench 16384 12 20 | 18182313 18371581 18336838 19381275 19738012 19620225 +7.0% On my box, huge pages have a significant perf impact when using a big hash size. They also speed up TT initialization big time: vanilla (s) huge pages (s) speed-up ======================================================================= time stockfish bench 16384 1 1 | 5.37 1.48 3.6x In practice, huge pages with auto-defrag may always be enabled in the system, in which case this patch has no effect. This depends on the values in /sys/kernel/mm/transparent_hugepage/enabled and /sys/kernel/mm/transparent_hugepage/defrag. closes https://github.com/official-stockfish/Stockfish/pull/2463 No functional change
Copyright 2011–2025 Next Chess Move LLC