Dev Builds » 20200810-1717

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 02:38:58 584979 1624 21 1038 565 -255.43 ± 9.99 233 551 28 0 0 -702.34 ± 67.37
ncm-dbt-02 02:38:27 586334 1596 37 1034 525 -254.55 ± 11.36 254 491 51 2 0 -578.98 ± 48.76
ncm-dbt-03 02:38:04 585920 1604 25 1058 521 -265.79 ± 11.07 264 506 31 1 0 -671.06 ± 63.65
ncm-dbt-04 02:39:09 568486 1620 28 1048 544 -257.38 ± 10.7 250 520 40 0 0 -638.63 ± 55.51
ncm-dbt-05 02:39:15 581811 1604 21 1021 562 -253.84 ± 10.1 228 545 28 1 0 -687.95 ± 67.28
8048 132 5199 2717 -257.36 ± 4.76 1229 2613 178 4 0 -650.41 ± 25.69

Test Detail

ID Host Base NPS Games WLD Standard Elo Ptnml(0-2) Gamepair Elo CLI PGN
461638 ncm-dbt-02 587240 96 1 65 30 -279.55 ± 49.7 18 28 2 0 0 -668.67 ± 342.59
461637 ncm-dbt-03 587325 104 3 71 30 -271.66 ± 47.97 19 30 3 0 0 -610.76 ± 361.79
461636 ncm-dbt-05 579992 104 1 68 35 -265.89 ± 40.27 16 35 1 0 0 -804.79 ± 292.87
461635 ncm-dbt-04 569669 120 3 79 38 -259.49 ± 42.57 20 36 4 0 0 -584.87 ± 318.28
461634 ncm-dbt-01 584874 124 2 80 42 -257.01 ± 36.51 18 42 2 0 0 -713.96 ± 319.61
461633 ncm-dbt-02 583279 500 13 316 171 -244.09 ± 19.27 72 159 19 0 0 -561.34 ± 83.33
461632 ncm-dbt-03 586012 500 11 322 167 -253.02 ± 20.18 78 156 15 1 0 -581.4 ± 95.42
461631 ncm-dbt-05 581070 500 5 302 193 -237.58 ± 16.07 57 183 10 0 0 -676.04 ± 124.55
461630 ncm-dbt-04 569629 500 7 319 174 -254.15 ± 18.24 72 168 10 0 0 -676.04 ± 124.55
461629 ncm-dbt-01 584495 500 8 327 165 -262.24 ± 18.7 77 165 8 0 0 -715.51 ± 146.53
461628 ncm-dbt-02 586858 500 13 321 166 -249.64 ± 19.99 76 157 16 1 0 -571.11 ± 91.87
461627 ncm-dbt-03 583698 500 4 331 165 -271.78 ± 18.67 81 165 4 0 0 -837.28 ± 240.94
461626 ncm-dbt-01 585000 500 5 315 180 -251.89 ± 17.4 68 174 8 0 0 -715.51 ± 146.53
461625 ncm-dbt-05 582903 500 7 322 171 -257.59 ± 18.12 73 169 8 0 0 -715.51 ± 146.53
461624 ncm-dbt-04 567561 500 9 318 173 -250.76 ± 18.88 73 163 14 0 0 -616.18 ± 100.06
461623 ncm-dbt-03 586646 500 7 334 159 -271.78 ± 20.15 86 155 9 0 0 -694.7 ± 134.18
461622 ncm-dbt-02 587962 500 10 332 158 -265.78 ± 21.55 88 147 14 1 0 -592.29 ± 99.42
461621 ncm-dbt-01 585548 500 6 316 178 -251.89 ± 17.95 70 170 10 0 0 -676.04 ± 124.55
461620 ncm-dbt-04 567086 500 9 332 159 -266.96 ± 20.39 85 153 12 0 0 -643.67 ± 110.29
461619 ncm-dbt-05 583279 500 8 329 163 -264.59 ± 20.09 82 158 9 1 0 -659.17 ± 131.36

Commit

Commit ID 21df37d7fd4dcc9b4a9c319382cc43685c0259c8
Author Fanael Linithien
Date 2020-08-10 17:17:57 UTC
Provide vectorized NNUE code for SSE2 and MMX targets This patch allows old x86 CPUs, from AMD K8 (which the x86-64 baseline targets) all the way down to the Pentium MMX, to benefit from NNUE with comparable performance hit versus hand-written eval as on more modern processors. NPS of the bench with NNUE enabled on a Pentium III 1.13 GHz (using the MMX code): master: 38951 this patch: 80586 NPS of the bench with NNUE enabled using baseline x86-64 arch, which is how linux distros are likely to package stockfish, on a modern CPU (using the SSE2 code): master: 882584 this patch: 1203945 closes https://github.com/official-stockfish/Stockfish/pull/2956 No functional change.
Copyright 2011–2026 Next Chess Move LLC