Dev Builds » 20230922-1715

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 09:54:14 1205353 3338 1445 281 1612 +126.46 ± 5.36 0 42 469 1110 48 +280.84 ± 15.74
ncm-dbt-02 09:50:27 1194682 3316 1409 277 1630 +123.56 ± 5.28 0 39 487 1093 39 +274.98 ± 15.44
ncm-dbt-03 09:55:05 1240985 3348 1472 257 1619 +132.1 ± 5.4 1 40 432 1145 56 +296.57 ± 16.41
ncm-dbt-04 09:55:11 1222387 3342 1467 273 1602 +129.86 ± 5.19 0 32 457 1138 44 +293.43 ± 15.94
ncm-dbt-05 09:51:30 1232715 3322 1423 267 1632 +126.17 ± 5.26 1 45 443 1141 31 +286.71 ± 16.2
ncm-dbt-06 09:55:03 1231962 3334 1442 294 1598 +124.73 ± 5.4 1 45 471 1105 45 +276.85 ± 15.71
20000 8658 1649 9693 +127.15 ± 2.17 3 243 2759 6732 263 +284.79 ± 6.48

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
204032 ncm-dbt-02 1196340 316 133 17 166 +133.78 ± 16.7 0 5 34 117 2 +316.45 ± 59.55
204031 ncm-dbt-05 1213011 322 131 19 172 +126.1 ± 16.7 0 5 41 113 2 +290.16 ± 54.16
204030 ncm-dbt-04 1225285 342 158 33 151 +133.14 ± 16.85 0 4 44 117 6 +298.56 ± 52.28
204029 ncm-dbt-06 1241944 334 146 16 172 +142.75 ± 16.81 0 4 35 122 6 +332.17 ± 58.89
204028 ncm-dbt-01 1214878 338 144 29 165 +123.11 ± 15.92 0 4 48 115 2 +280.82 ± 49.92
204027 ncm-dbt-03 1250555 348 146 38 164 +111.5 ± 18.52 0 9 55 103 7 +230.4 ± 46.41
204026 ncm-dbt-02 1196625 500 215 47 238 +121.45 ± 13.14 0 4 79 162 5 +270.57 ± 38.47
204025 ncm-dbt-05 1236958 500 207 47 246 +115.22 ± 13.55 0 8 77 162 3 +256.44 ± 39.15
204024 ncm-dbt-04 1212418 500 224 37 239 +136.56 ± 12.18 0 2 63 181 4 +324.17 ± 43.32
204023 ncm-dbt-06 1219823 500 225 40 235 +134.95 ± 14.65 1 4 66 167 12 +298.62 ± 42.41
204022 ncm-dbt-03 1228035 500 223 40 237 +133.34 ± 13.63 0 8 56 181 5 +309.64 ± 45.98
204021 ncm-dbt-01 1190153 500 225 39 236 +135.76 ± 13.57 0 4 65 172 9 +306.84 ± 42.73
204020 ncm-dbt-05 1245637 500 212 40 248 +124.6 ± 13.27 0 7 67 173 3 +285.49 ± 42.08
204019 ncm-dbt-02 1196270 500 207 47 246 +115.22 ± 14.53 0 7 86 147 10 +240.82 ± 36.9
204018 ncm-dbt-06 1227151 500 218 42 240 +127.76 ± 12.84 0 4 70 172 4 +293.29 ± 41.07
204017 ncm-dbt-04 1211601 500 215 43 242 +124.6 ± 13.97 0 6 74 162 8 +273.0 ± 39.95
204016 ncm-dbt-03 1261469 500 212 36 252 +127.76 ± 14.27 1 6 66 170 7 +288.06 ± 42.4
204015 ncm-dbt-01 1210997 500 222 35 243 +136.56 ± 12.98 0 6 55 185 4 +324.17 ± 46.56
204014 ncm-dbt-02 1193823 500 216 62 222 +110.6 ± 14.69 0 13 75 157 5 +238.66 ± 39.57
204013 ncm-dbt-05 1235276 500 218 40 242 +129.35 ± 13.36 0 5 68 171 6 +293.29 ± 41.75
204012 ncm-dbt-06 1231631 500 221 55 224 +119.89 ± 13.85 0 6 79 158 7 +261.07 ± 38.58
204011 ncm-dbt-04 1224019 500 220 47 233 +125.38 ± 12.7 0 3 75 168 4 +285.49 ± 39.5
204010 ncm-dbt-01 1201766 500 215 42 243 +125.38 ± 13.96 0 7 70 166 7 +277.93 ± 41.14
204009 ncm-dbt-03 1237926 500 217 39 244 +129.35 ± 13.36 0 5 68 171 6 +293.29 ± 41.75
204008 ncm-dbt-02 1191787 500 212 37 251 +126.97 ± 12.66 0 3 73 170 4 +290.66 ± 40.08
204007 ncm-dbt-05 1225592 500 226 40 234 +135.76 ± 14.12 0 7 59 175 9 +306.84 ± 44.88
204006 ncm-dbt-06 1223014 500 217 53 230 +118.33 ± 14.2 0 9 74 161 6 +258.75 ± 39.96
204005 ncm-dbt-01 1222936 500 215 47 238 +121.45 ± 14.01 0 6 78 158 8 +263.42 ± 38.85
204004 ncm-dbt-04 1247457 500 216 33 251 +133.34 ± 13.99 0 7 61 174 8 +301.33 ± 44.13
204003 ncm-dbt-03 1254317 500 226 35 239 +139.81 ± 14.19 0 4 64 169 13 +309.64 ± 43.08
204002 ncm-dbt-02 1193523 500 220 36 244 +134.15 ± 13.61 0 4 67 170 9 +301.33 ± 42.04
204001 ncm-dbt-06 1251327 500 202 42 256 +115.22 ± 14.37 0 11 73 161 5 +251.89 ± 40.18
204000 ncm-dbt-05 1226004 500 215 48 237 +120.67 ± 13.5 1 5 73 168 3 +275.45 ± 40.24
203999 ncm-dbt-01 1213067 500 208 39 253 +122.24 ± 14.5 0 9 71 162 8 +265.78 ± 40.81
203998 ncm-dbt-04 1220538 500 221 37 242 +134.15 ± 14.15 0 5 67 167 11 +295.94 ± 42.08
203997 ncm-dbt-03 1243310 500 224 42 234 +132.54 ± 13.1 0 5 63 177 5 +306.84 ± 43.45
203996 ncm-dbt-02 1194409 500 206 31 263 +126.97 ± 12.66 0 3 73 170 4 +290.66 ± 40.08
203995 ncm-dbt-01 1183675 500 216 50 234 +119.89 ± 14.35 0 6 82 152 10 +254.16 ± 37.82
203994 ncm-dbt-05 1246527 500 214 33 253 +131.74 ± 13.67 0 8 58 179 5 +304.07 ± 45.19
203993 ncm-dbt-03 1211287 500 224 27 249 +144.71 ± 13.86 0 3 60 174 13 +327.18 ± 44.54
203992 ncm-dbt-06 1228848 500 213 46 241 +120.67 ± 13.67 0 7 74 164 5 +268.17 ± 39.97
203991 ncm-dbt-04 1215397 500 213 43 244 +123.02 ± 12.93 0 5 73 169 3 +280.42 ± 40.21

Commit

Commit ID 95fe2b9a9d33811a7fcad1cdfea79c54e8fdb074
Author mstembera
Date 2023-09-22 17:15:34 UTC
Reduce SIMD register count from 32 to 16 in the case of avx512 and vnni512 archs. Up to 17% speedup, depending on the compiler, e.g. ``` AMD pro 7840u (zen4 phoenix apu 4nm) bash bench_parallel.sh ./stockfish_avx512_gcc13 ./stockfish_avx512_pr_gcc13 20 10 sf_base = 1077737 +/- 8446 (95%) sf_test = 1264268 +/- 8543 (95%) diff = 186531 +/- 4280 (95%) speedup = 17.308% +/- 0.397% (95%) ``` Prior to this patch, it appears gcc spills registers. closes https://github.com/official-stockfish/Stockfish/pull/4796 No functional change
Copyright 2011–2024 Next Chess Move LLC