Dev Builds » 20230614-1636

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 06:51:39 582252 4000 1340 704 1956 +55.71 ± 5.07 1 194 981 816 8 +113.1 ± 10.83
ncm-dbt-02 06:53:35 586181 4014 1370 703 1941 +58.27 ± 5.07 1 182 987 823 14 +117.5 ± 10.79
ncm-dbt-03 06:53:04 585198 4000 1329 710 1961 +54.2 ± 5.01 2 190 998 807 3 +110.98 ± 10.72
ncm-dbt-04 06:51:43 570214 4000 1351 709 1940 +56.25 ± 5.08 0 193 983 813 11 +113.49 ± 10.82
ncm-dbt-05 06:53:15 584716 3986 1310 664 2012 +56.81 ± 5.09 3 189 965 831 5 +116.44 ± 10.93
20000 6700 3490 9810 +56.25 ± 2.26 7 948 4914 4090 41 +114.3 ± 4.84

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
408507 ncm-dbt-02 586520 14 4 3 7 +24.85 ± 111.54 0 2 2 3 0 +49.98 ± 262.59 ↓
408506 ncm-dbt-05 584034 486 157 86 243 +51.12 ± 14.63 1 24 121 97 0 +106.13 ± 30.93 ↓
408505 ncm-dbt-03 584537 500 180 104 216 +53.22 ± 13.73 0 21 133 95 1 +107.54 ± 29.18 ↓
408504 ncm-dbt-01 583866 500 176 84 240 +64.66 ± 13.39 0 15 129 105 1 +132.54 ± 29.53 ↓
408503 ncm-dbt-04 571150 500 161 77 262 +58.93 ± 14.69 0 26 115 108 1 +119.89 ± 31.85 ↓
408502 ncm-dbt-02 585843 500 152 69 279 +58.21 ± 14.79 0 26 117 105 2 +116.78 ± 31.56 ↓
408501 ncm-dbt-05 585885 500 160 75 265 +59.64 ± 14.04 0 22 121 107 0 +123.02 ± 30.92 ↓
408500 ncm-dbt-03 585590 500 163 99 238 +44.72 ± 14.37 1 27 129 93 0 +92.46 ± 29.89 ↓
408499 ncm-dbt-01 582444 500 175 93 232 +57.5 ± 14.08 0 23 122 105 0 +118.33 ± 30.8 ↓
408498 ncm-dbt-04 569190 500 168 104 228 +44.72 ± 13.26 0 21 145 83 1 +89.48 ± 27.53 ↓
408497 ncm-dbt-02 587240 500 167 88 245 +55.36 ± 14.26 0 23 127 98 2 +110.6 ± 30.08 ↓
408496 ncm-dbt-05 587155 500 170 82 248 +61.79 ± 14.82 0 25 114 109 2 +124.6 ± 31.99 ↓
408495 ncm-dbt-04 568474 500 166 78 256 +61.79 ± 13.99 0 18 129 100 3 +123.02 ± 29.65 ↓
408494 ncm-dbt-03 588430 500 162 76 262 +60.36 ± 14.89 0 27 111 111 1 +123.02 ± 32.44 ↓
408493 ncm-dbt-01 583196 500 159 81 260 +54.64 ± 14.08 0 23 127 99 1 +110.6 ± 30.08 ↓
408492 ncm-dbt-02 582778 500 186 94 220 +64.66 ± 13.97 0 17 127 103 3 +129.35 ± 29.91 ↓
408491 ncm-dbt-05 585042 500 166 92 242 +51.8 ± 14.35 0 27 122 101 0 +106.01 ± 30.85 ↓
408490 ncm-dbt-04 570989 500 173 85 242 +61.79 ± 13.85 0 18 128 102 2 +124.6 ± 29.8 ↓
408489 ncm-dbt-01 581818 500 173 87 240 +60.36 ± 14.07 0 21 123 105 1 +123.02 ± 30.61 ↓
408488 ncm-dbt-02 584537 500 185 90 225 +66.82 ± 14.48 0 20 118 109 3 +134.15 ± 31.33 ↓
408487 ncm-dbt-03 584622 500 152 83 265 +48.25 ± 14.16 0 27 127 96 0 +98.44 ± 30.15 ↓
408486 ncm-dbt-05 581610 500 166 94 240 +50.38 ± 15.45 0 34 112 102 2 +99.95 ± 32.27 ↓
408485 ncm-dbt-04 570589 500 175 80 245 +66.82 ± 14.89 0 25 106 118 1 +137.37 ± 33.21 ↓
408484 ncm-dbt-01 582903 500 172 87 241 +59.64 ± 14.32 0 24 117 109 0 +123.02 ± 31.54 ↓
408483 ncm-dbt-02 586266 500 156 95 249 +42.6 ± 15.16 1 32 124 91 2 +85.04 ± 30.63 ↓
408482 ncm-dbt-03 585759 500 175 90 235 +59.64 ± 14.18 1 20 122 107 0 +124.6 ± 30.76 ↓
408481 ncm-dbt-05 582736 500 165 79 256 +60.36 ± 13.79 0 19 127 103 1 +123.02 ± 29.98 ↓
408480 ncm-dbt-04 569709 500 159 99 242 +41.89 ± 14.99 0 35 121 93 1 +83.57 ± 31.04 ↓
408479 ncm-dbt-01 577028 500 164 86 250 +54.64 ± 14.22 0 24 125 100 1 +110.6 ± 30.39 ↓
408478 ncm-dbt-02 586774 500 182 88 230 +66.1 ± 13.6 0 16 125 108 1 +135.76 ± 30.17 ↓
408477 ncm-dbt-03 583572 500 164 81 255 +58.21 ± 13.55 0 19 129 102 0 +119.89 ± 29.69 ↓
408476 ncm-dbt-05 585000 500 150 81 269 +48.26 ± 14.3 2 22 131 95 0 +101.46 ± 29.54 ↓
408475 ncm-dbt-04 570829 500 179 90 231 +62.51 ± 14.31 0 23 115 112 0 +129.35 ± 31.83 ↓
408474 ncm-dbt-01 583112 500 153 96 251 +39.78 ± 15.62 1 37 118 92 2 +79.17 ± 31.44 ↓
408473 ncm-dbt-02 586901 500 171 95 234 +53.22 ± 13.59 0 20 135 94 1 +107.54 ± 28.87 ↓
408472 ncm-dbt-03 584076 500 166 94 240 +50.38 ± 14.28 0 26 127 96 1 +101.46 ± 30.14 ↓
408471 ncm-dbt-05 586266 500 176 75 249 +71.16 ± 13.64 0 16 117 117 0 +148.85 ± 31.38 ↓
408470 ncm-dbt-01 583656 500 168 90 242 +54.64 ± 14.76 0 27 120 101 2 +109.07 ± 31.14 ↓
408469 ncm-dbt-04 570789 500 170 96 234 +51.8 ± 14.62 0 27 124 97 2 +102.97 ± 30.57 ↓
408468 ncm-dbt-02 588771 500 167 81 252 +60.36 ± 14.62 0 26 112 112 0 +124.6 ± 32.29 ↓
408467 ncm-dbt-03 585000 500 167 83 250 +58.93 ± 14.15 0 23 120 107 0 +121.46 ± 31.09 ↓

Commit

Commit ID 32d3284df5b2fd395504efa5319d64856902fef1
Author AndrovT
Date 2023-06-14 16:36:39 UTC
Permute master net weights to increase sparsity Activation data collection using https://github.com/AndrovT/Stockfish/commit/ac468039ab544b03ad9a22c859a4217729c10a77 run as bench 16 1 13 varied_1000.epd depth NNUE log.bin on FENs from https://gist.github.com/AndrovT/7eae6918eb50764227e2bafe7938953c. Permutation found using https://gist.github.com/AndrovT/359c831b7223c637e9156b01eb96949e. Uses a greedy algorithm that goes sequentially through the output positions and chooses a neuron for that position such that the number of nonzero quartets is the smallest. Net weights permuted using https://gist.github.com/AndrovT/9e3fbaebb7082734dc84d27e02094cb3. Benchmark: Result of 100 runs of 'bench 16 1 13 default depth NNUE' ======================================================== base (...kfish-master) = 885869 +/- 7395 test (./stockfish ) = 895885 +/- 7368 diff = +10016 +/- 2984 speedup = +0.0113 P(speedup > 0) = 1.0000 Passed STC: https://tests.stockfishchess.org/tests/view/648866c4713491385c804728 LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 126784 W: 34003 L: 33586 D: 59195 Ptnml(0-2): 283, 13001, 36437, 13358, 313 closes https://github.com/official-stockfish/Stockfish/pull/4620 No functional change.
Copyright 2011–2025 Next Chess Move LLC
Feedback & Support Terms of Service Privacy Policy