Dev Builds » 20260110-1446

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 06:44:55 582179 3998 1749 370 1879 +124.96 ± 4.52 0 29 592 1348 30 +284.75 ± 13.96
ncm-dbt-02 06:43:10 585840 4000 1737 350 1913 +125.68 ± 4.56 0 37 567 1368 28 +287.74 ± 14.3
ncm-dbt-03 06:43:50 585998 4002 1743 392 1867 +122.07 ± 4.52 0 32 611 1333 25 +277.1 ± 13.74
ncm-dbt-04 06:43:43 565985 4000 1743 335 1922 +127.76 ± 4.51 0 29 566 1373 32 +293.29 ± 14.29
ncm-dbt-05 06:44:46 586772 4000 1721 370 1909 +122.14 ± 4.57 0 37 602 1334 27 +276.69 ± 13.86
20000 8693 1817 9490 +124.52 ± 2.03 0 164 2938 6756 142 +283.83 ± 6.27

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
449679 ncm-dbt-03 583531 2 1 0 1 +189.7 ± 55.98 0 0 0 1 0 +1129.65 ± 376.02
449678 ncm-dbt-05 582945 500 221 41 238 +130.94 ± 13.32 0 6 63 176 5 +301.33 ± 43.45
449677 ncm-dbt-01 583489 498 221 42 235 +130.72 ± 13.18 0 4 68 171 6 +297.8 ± 41.71
449676 ncm-dbt-02 586901 500 211 40 249 +123.81 ± 13.1 0 7 67 174 2 +285.49 ± 42.08
449675 ncm-dbt-04 565744 500 226 36 238 +138.99 ± 13.67 0 5 59 177 9 +318.25 ± 44.96
449674 ncm-dbt-03 585801 500 220 50 230 +123.02 ± 13.29 0 7 69 171 3 +280.42 ± 41.44
449673 ncm-dbt-01 581361 500 222 43 235 +130.14 ± 12.39 0 2 71 173 4 +301.33 ± 40.6
449672 ncm-dbt-05 580033 500 209 42 249 +120.67 ± 12.97 0 4 79 163 4 +270.57 ± 38.47
449671 ncm-dbt-04 567126 500 207 37 256 +123.02 ± 12.37 0 2 79 166 3 +280.42 ± 38.28
449670 ncm-dbt-03 586646 500 221 45 234 +127.76 ± 12.45 0 3 71 173 3 +295.94 ± 40.69
449669 ncm-dbt-02 586266 500 217 32 251 +134.95 ± 12.44 0 5 57 186 2 +324.17 ± 45.77
449668 ncm-dbt-01 581569 500 216 48 236 +121.45 ± 12.95 0 6 72 170 2 +277.93 ± 40.53
449667 ncm-dbt-05 580904 500 209 59 232 +107.54 ± 13.4 0 7 89 151 3 +234.38 ± 36.22
449666 ncm-dbt-04 565272 500 224 44 232 +130.94 ± 11.76 0 2 67 180 1 +312.48 ± 41.9
449665 ncm-dbt-02 585421 500 217 46 237 +123.81 ± 13.81 0 8 68 169 5 +277.93 ± 41.74
449664 ncm-dbt-03 587749 500 221 39 240 +132.54 ± 12.32 0 2 68 176 4 +309.64 ± 41.56
449663 ncm-dbt-01 583447 500 214 47 239 +120.67 ± 13.32 0 5 78 162 5 +268.17 ± 38.8
449662 ncm-dbt-04 562534 500 220 48 232 +124.6 ± 12.9 0 4 74 168 4 +282.94 ± 39.86
449661 ncm-dbt-05 580075 500 212 55 233 +112.91 ± 12.86 0 5 85 158 2 +251.89 ± 37.03
449660 ncm-dbt-02 585928 500 220 47 233 +125.38 ± 12.7 0 5 69 174 2 +290.66 ± 41.43
449659 ncm-dbt-03 583866 500 217 49 234 +121.45 ± 12.77 0 4 77 166 3 +275.45 ± 39.01
449658 ncm-dbt-01 583489 500 218 53 229 +119.11 ± 13.17 0 6 76 165 3 +268.17 ± 39.39
449657 ncm-dbt-04 568156 500 215 40 245 +126.97 ± 11.88 0 1 75 172 2 +295.94 ± 39.27
449656 ncm-dbt-05 582485 500 218 45 237 +125.38 ± 12.51 0 1 80 164 5 +282.94 ± 37.88
449655 ncm-dbt-02 587113 500 221 52 227 +122.24 ± 12.94 0 3 80 162 5 +273.0 ± 38.12
449654 ncm-dbt-03 584706 500 220 56 224 +118.33 ± 12.63 0 3 83 161 3 +265.78 ± 37.35
449653 ncm-dbt-01 579702 500 223 52 225 +123.81 ± 12.73 0 2 80 163 5 +277.93 ± 38.02
449652 ncm-dbt-02 583363 500 218 42 240 +127.76 ± 13.02 0 3 74 167 6 +288.06 ± 39.79
449651 ncm-dbt-03 588515 500 214 52 234 +116.77 ± 13.36 0 5 83 157 5 +256.44 ± 37.51
449650 ncm-dbt-05 581652 500 213 44 243 +122.24 ± 12.76 0 5 73 170 2 +280.42 ± 40.21
449649 ncm-dbt-04 567958 500 219 39 242 +130.94 ± 13.32 0 5 66 173 6 +298.62 ± 42.41
449648 ncm-dbt-03 585042 500 220 52 228 +121.45 ± 12.58 0 4 76 168 2 +277.93 ± 39.29
449647 ncm-dbt-01 578918 500 215 38 247 +128.55 ± 12.24 0 2 72 173 3 +298.62 ± 40.29
449646 ncm-dbt-04 566138 500 216 52 232 +118.33 ± 13.86 0 9 72 165 4 +263.42 ± 40.52
449645 ncm-dbt-02 586181 500 210 54 236 +112.14 ± 12.87 0 3 92 151 4 +245.2 ± 35.25
449644 ncm-dbt-05 583405 500 221 37 242 +134.15 ± 13.24 0 6 59 180 5 +312.48 ± 44.93
449643 ncm-dbt-02 585548 500 223 37 240 +135.76 ± 12.01 0 3 60 185 2 +327.18 ± 44.54
449642 ncm-dbt-03 588132 500 209 49 242 +115.22 ± 12.67 0 4 84 160 2 +258.75 ± 37.2
449641 ncm-dbt-01 585464 500 220 47 233 +125.38 ± 12.12 0 2 75 171 2 +290.66 ± 39.4
449640 ncm-dbt-04 564958 500 216 39 245 +128.55 ± 12.04 0 1 74 172 3 +298.62 ± 39.56
449639 ncm-dbt-05 622683 500 218 47 235 +123.81 ± 12.16 0 3 74 172 1 +288.06 ± 39.79

Commit

Commit ID b4d4eecfb24633c89d0f7fde3f76a3c7c4e0dac1
Author anematode
Date 2026-01-10 14:46:01 UTC
Make shared history allocation aware of non-uniform cache access Although shared history has been successful overall, it led to some speed issues with large numbers of threads. Originally we just split by NUMA node, but on systems with non-unified L3 caches (most AMD workstation and server CPUs, and some Intel E-core based server CPUs), this can still lead to a speed penalty at the default config. Thus, we decided to further subdivide the shared history based on the L3 cache structure. Based on this test, the original SPRTs, and speed experiments, we decided that grouping L3 domains to reach 32 threads per SharedHistories was a reasonable balance for affected systems – but we may revisit this in the future. See the PR for full details. In an extreme case, a single-socket EPYC 9755 configured with 1 numa domain per socket, the nps increases from: Nodes/second : 182827480 to Nodes/second : 229118365 In many cases, when L3 caches are shared between many threads, or when several numa nodes are already configured per socket, this patch does not influence the default. This default setting can adjusted with the existing NumaPolicy option. closes https://github.com/official-stockfish/Stockfish/pull/6526 No functional change.
Copyright 2011–2026 Next Chess Move LLC