Dev Builds » 20230822-0914

You are viewing an old NCM Stockfish dev build test. You may find the most recent dev build tests using Stockfish 15 as the baseline here.

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 14. 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 09:52:12 1141158 3318 1390 284 1644 +120.41 +/- 5.35 0 51 484 1091 33 +267.43 +/- 15.5
ncm-dbt-02 09:49:23 1194677 3316 1407 292 1617 +121.55 +/- 5.33 0 41 502 1074 41 +268.03 +/- 15.2
ncm-dbt-03 09:53:04 1237191 3348 1446 293 1609 +124.75 +/- 5.34 0 52 453 1133 36 +279.96 +/- 16.02
ncm-dbt-04 09:53:30 1223313 3358 1453 294 1611 +125.05 +/- 5.38 1 47 467 1120 44 +278.35 +/- 15.78
ncm-dbt-05 09:49:50 1227778 3314 1421 260 1633 +127.1 +/- 5.12 0 31 468 1124 34 +288.16 +/- 15.74
ncm-dbt-06 09:53:45 1232869 3346 1437 276 1633 +125.77 +/- 5.24 2 38 465 1133 35 +284.37 +/- 15.81
20000 8554 1699 9747 +124.11 +/- 2.16 3 260 2839 6675 223 +277.62 +/- 6.39

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
200728 ncm-dbt-05 1235692 314 136 21 157 +133.44 +/- 17.51 0 3 42 106 6 +297.46 +/- 53.53
200727 ncm-dbt-01 1148504 318 134 34 150 +113.08 +/- 18.6 0 8 47 100 4 +242.88 +/- 50.27
200726 ncm-dbt-02 1195080 316 134 28 154 +121.23 +/- 16.64 0 2 52 100 4 +266.71 +/- 47.48
200725 ncm-dbt-06 1250681 346 144 23 179 +126.85 +/- 14.95 0 3 47 122 1 +297.04 +/- 50.46
200724 ncm-dbt-03 1240788 348 146 31 171 +119.29 +/- 17.18 0 8 46 117 3 +265.59 +/- 50.76
200723 ncm-dbt-04 1245893 358 155 40 163 +115.7 +/- 18.41 0 9 54 108 8 +239.61 +/- 46.83
200722 ncm-dbt-05 1214289 500 220 39 241 +131.74 +/- 13.85 0 5 68 168 9 +293.29 +/- 41.75
200721 ncm-dbt-01 1128660 500 205 47 248 +113.68 +/- 13.72 0 5 89 149 7 +243.0 +/- 36.1
200720 ncm-dbt-02 1192333 500 216 38 246 +129.35 +/- 13.72 0 5 70 167 8 +288.06 +/- 41.11
200719 ncm-dbt-06 1208329 500 209 51 240 +113.68 +/- 14.22 1 8 77 160 4 +251.89 +/- 39.16
200718 ncm-dbt-03 1248142 500 203 52 245 +108.3 +/- 14.05 0 10 83 153 4 +234.38 +/- 37.67
200717 ncm-dbt-04 1193511 500 218 41 241 +128.55 +/- 13.91 0 8 63 173 6 +290.66 +/- 43.37
200716 ncm-dbt-01 1143616 500 215 41 244 +126.17 +/- 13.24 0 4 74 166 6 +282.94 +/- 39.86
200715 ncm-dbt-05 1226137 500 206 42 252 +118.33 +/- 13.35 0 4 84 156 6 +258.75 +/- 37.2
200714 ncm-dbt-02 1192767 500 210 49 241 +116.0 +/- 13.54 0 5 85 154 6 +251.89 +/- 37.03
200713 ncm-dbt-06 1239473 500 214 37 249 +128.55 +/- 13.19 0 5 68 172 5 +293.29 +/- 41.75
200712 ncm-dbt-04 1212897 500 213 49 238 +118.33 +/- 14.36 1 9 69 167 4 +265.78 +/- 41.36
200711 ncm-dbt-03 1221780 500 225 38 237 +136.56 +/- 13.73 0 5 62 174 9 +309.64 +/- 43.82
200710 ncm-dbt-01 1148027 500 212 38 250 +126.17 +/- 13.6 0 8 64 174 4 +288.06 +/- 43.03
200709 ncm-dbt-05 1237403 500 215 34 251 +131.74 +/- 12.14 0 4 62 183 1 +315.35 +/- 43.81
200708 ncm-dbt-02 1197900 500 206 41 253 +119.11 +/- 14.36 0 9 74 160 7 +258.75 +/- 39.96
200707 ncm-dbt-06 1230003 500 214 38 248 +127.76 +/- 12.26 0 3 70 175 2 +298.62 +/- 41.01
200706 ncm-dbt-03 1244717 500 212 47 241 +119.11 +/- 13.52 0 8 72 167 3 +268.17 +/- 40.54
200705 ncm-dbt-04 1222741 500 214 46 240 +121.45 +/- 13.84 0 7 74 163 6 +268.17 +/- 39.97
200704 ncm-dbt-01 1136530 500 207 41 252 +119.89 +/- 14.51 0 11 68 165 6 +263.42 +/- 41.6
200703 ncm-dbt-02 1196872 500 217 45 238 +124.6 +/- 13.62 0 7 69 169 5 +280.42 +/- 41.44
200702 ncm-dbt-05 1224615 500 211 41 248 +123.02 +/- 12.93 0 4 76 166 4 +277.93 +/- 39.29
200701 ncm-dbt-06 1242463 500 217 39 244 +129.35 +/- 13.72 0 7 64 173 6 +293.29 +/- 43.07
200700 ncm-dbt-03 1238471 500 213 38 249 +126.97 +/- 14.11 0 9 63 172 6 +285.49 +/- 43.31
200699 ncm-dbt-04 1224118 500 218 42 240 +127.76 +/- 14.44 0 5 76 157 12 +273.0 +/- 39.35
200698 ncm-dbt-01 1142349 500 206 45 249 +116.0 +/- 14.05 0 11 70 166 3 +258.75 +/- 41.02
200697 ncm-dbt-02 1192329 500 202 40 258 +116.77 +/- 13.88 0 7 80 157 6 +254.16 +/- 38.36
200696 ncm-dbt-05 1236181 500 217 41 242 +127.76 +/- 12.84 0 3 73 169 5 +290.66 +/- 40.08
200695 ncm-dbt-03 1239708 500 228 46 226 +132.54 +/- 12.91 0 5 62 179 4 +309.64 +/- 43.82
200694 ncm-dbt-06 1220610 500 219 43 238 +127.76 +/- 14.44 1 6 67 168 8 +285.49 +/- 42.08
200693 ncm-dbt-04 1212290 500 225 37 238 +137.37 +/- 12.76 0 3 62 179 6 +321.19 +/- 43.77
200692 ncm-dbt-01 1140422 500 211 38 251 +125.38 +/- 12.7 0 4 72 171 3 +288.06 +/- 40.46
200691 ncm-dbt-05 1220134 500 216 42 242 +126.17 +/- 13.42 0 8 63 176 3 +290.66 +/- 43.37
200690 ncm-dbt-03 1226731 500 219 41 240 +129.35 +/- 13.9 0 7 65 171 7 +290.66 +/- 42.73
200689 ncm-dbt-02 1195460 500 222 51 227 +123.81 +/- 13.46 0 6 72 167 5 +277.93 +/- 40.53
200688 ncm-dbt-06 1238524 500 220 45 235 +126.97 +/- 14.11 0 6 72 163 9 +277.93 +/- 40.53
200687 ncm-dbt-04 1251743 500 210 39 251 +123.81 +/- 12.92 0 6 69 173 2 +285.49 +/- 41.44

Commit

Commit ID c6f62363a657263a567a0cc9bae09f3c4016156d
Author Gian-Carlo Pascutto
Date 2023-08-22 09:14:19 UTC
Simplify Square Clipped ReLU code. Squared numbers are never negative, so barring any wraparound there is no need to clamp to 0. From reading the code, there's no obvious way to get wraparound, so the entire operation can be simplified away. Updated original truncated code comments to be sensible. Verified by running ./stockfish bench 128 1 24 and by the following test: STC: https://tests.stockfishchess.org/tests/view/64da4db95b17f7c21c0eabe7 LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 60224 W: 15425 L: 15236 D: 29563 Ptnml(0-2): 195, 6576, 16382, 6763, 196 closes https://github.com/official-stockfish/Stockfish/pull/4751 No functional change
Copyright 2011–2024 Next Chess Move LLC