Dev Builds » 20200806-1437

Use this dev build

NCM plays each Stockfish dev build 20,000 times against Stockfish 7. This yields an approximate Elo difference and establishes confidence in the strength of the dev builds.

Summary

Host Duration Avg Base NPS Games Wins Losses Draws Elo
ncm-et-3 08:46:30 1963248 3400 2236 99 1065 +256.75 +/- 10.43
ncm-et-4 08:46:20 1967486 3406 2227 100 1079 +254.44 +/- 10.36
ncm-et-9 08:46:38 1964870 3386 2225 73 1088 +260.82 +/- 10.3
ncm-et-10 08:46:35 1963223 3394 2304 90 1000 +270.77 +/- 10.77
ncm-et-13 07:48:17 1954702 3000 1967 71 962 +258.75 +/- 10.96
ncm-et-15 08:46:41 1967795 3414 2229 97 1088 +254.44 +/- 10.32
20000 13188 530 6282 +259.27 +/- 4.29

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
138740 ncm-et-9 2020-09-17 16:26 01:01:05 1966169 386 263 7 116 +277.44 +/- 31.73 ↓
138739 ncm-et-10 2020-09-17 16:25 01:01:20 1965243 394 257 10 127 +255.82 +/- 30.32 ↓
138738 ncm-et-3 2020-09-17 16:25 01:02:09 1963378 400 271 16 113 +261.95 +/- 32.2 ↓
138737 ncm-et-4 2020-09-17 16:25 01:02:12 1962457 406 267 8 131 +262.2 +/- 29.79 ↓
138736 ncm-et-15 2020-09-17 16:23 01:03:26 1962303 414 270 13 131 +252.33 +/- 29.89 ↓
138735 ncm-et-13 2020-09-17 15:10 01:17:23 1959240 500 325 16 159 +250.77 +/- 27.11 ↓
138734 ncm-et-3 2020-09-17 15:07 01:16:13 1962149 500 327 13 160 +256.44 +/- 26.99 ↓
138733 ncm-et-9 2020-09-17 15:07 01:17:31 1964006 500 321 14 165 +248.53 +/- 26.57 ↓
138732 ncm-et-4 2020-09-17 15:07 01:16:40 1971726 500 332 9 159 +266.97 +/- 27.01 ↓
138731 ncm-et-10 2020-09-17 15:06 01:17:57 1958316 500 355 12 133 +291.97 +/- 29.7 ↓
138730 ncm-et-15 2020-09-17 15:05 01:17:11 1965231 500 316 13 171 +244.1 +/- 26.05 ↓
138729 ncm-et-13 2020-09-17 13:51 01:17:51 1963532 500 343 8 149 +281.68 +/- 27.94 ↓
138728 ncm-et-4 2020-09-17 13:49 01:17:11 1966459 500 327 18 155 +250.77 +/- 27.47 ↓
138727 ncm-et-9 2020-09-17 13:49 01:17:35 1961861 500 324 14 162 +251.89 +/- 26.83 ↓
138726 ncm-et-3 2020-09-17 13:48 01:18:18 1963379 500 335 13 152 +265.78 +/- 27.73 ↓
138725 ncm-et-15 2020-09-17 13:47 01:16:50 1974980 500 327 13 160 +256.44 +/- 26.99 ↓
138724 ncm-et-10 2020-09-17 13:47 01:18:35 1956486 500 335 10 155 +269.37 +/- 27.4 ↓
138723 ncm-et-13 2020-09-17 12:34 01:16:26 1960614 500 333 9 158 +268.17 +/- 27.1 ↓
138722 ncm-et-4 2020-09-17 12:31 01:16:49 1970664 500 320 22 158 +238.66 +/- 27.2 ↓
138721 ncm-et-3 2020-09-17 12:30 01:17:14 1962303 500 326 19 155 +248.53 +/- 27.47 ↓
138720 ncm-et-9 2020-09-17 12:30 01:17:58 1964789 500 331 10 159 +264.6 +/- 27.03 ↓
138719 ncm-et-10 2020-09-17 12:29 01:16:43 1961698 500 347 12 141 +281.68 +/- 28.82 ↓
138718 ncm-et-15 2020-09-17 12:29 01:17:46 1965232 500 329 14 157 +257.59 +/- 27.27 ↓
138717 ncm-et-13 2020-09-17 11:14 01:18:27 1926048 500 324 14 162 +251.89 +/- 26.83 ↓
138716 ncm-et-4 2020-09-17 11:12 01:17:21 1963690 500 320 17 163 +244.1 +/- 26.77 ↓
138715 ncm-et-9 2020-09-17 11:12 01:16:23 1959694 500 335 9 156 +270.57 +/- 27.29 ↓
138714 ncm-et-15 2020-09-17 11:11 01:16:08 1972030 500 334 13 153 +264.6 +/- 27.63 ↓
138713 ncm-et-3 2020-09-17 11:11 01:17:53 1963686 500 331 15 154 +258.75 +/- 27.55 ↓
138712 ncm-et-10 2020-09-17 11:10 01:17:35 1967099 500 329 18 153 +253.02 +/- 27.65 ↓
138711 ncm-et-13 2020-09-17 09:56 01:17:42 1960006 500 335 12 153 +266.97 +/- 27.62 ↓
138710 ncm-et-4 2020-09-17 09:54 01:17:11 1971571 500 319 13 168 +247.41 +/- 26.3 ↓
138709 ncm-et-9 2020-09-17 09:54 01:17:28 1970331 500 329 14 157 +257.59 +/- 27.27 ↓
138708 ncm-et-15 2020-09-17 09:53 01:17:36 1970018 500 336 13 151 +266.97 +/- 27.82 ↓
138707 ncm-et-3 2020-09-17 09:52 01:17:23 1963227 500 306 14 180 +232.26 +/- 25.36 ↓
138706 ncm-et-10 2020-09-17 09:52 01:17:10 1963862 500 336 14 150 +265.78 +/- 27.93 ↓
138705 ncm-et-13 2020-09-17 08:34 01:20:28 1958776 500 307 12 181 +235.45 +/- 25.24 ↓
138704 ncm-et-4 2020-09-17 08:34 01:18:56 1965841 500 342 13 145 +274.22 +/- 28.41 ↓
138703 ncm-et-3 2020-09-17 08:34 01:17:20 1964617 500 340 9 151 +276.69 +/- 27.77 ↓
138702 ncm-et-15 2020-09-17 08:34 01:17:44 1964771 500 317 18 165 +239.74 +/- 26.6 ↓
138701 ncm-et-9 2020-09-17 08:34 01:18:38 1967243 500 322 5 173 +259.91 +/- 25.67 ↓
138700 ncm-et-10 2020-09-17 08:34 01:17:15 1969858 500 345 14 141 +276.69 +/- 28.83 ↓

Commit

Commit ID 84f3e867903f62480c33243dd0ecbffd342796fc
Author nodchip
Date 2020-08-06 14:37:45 UTC
Add NNUE evaluation This patch ports the efficiently updatable neural network (NNUE) evaluation to Stockfish. Both the NNUE and the classical evaluations are available, and can be used to assign a value to a position that is later used in alpha-beta (PVS) search to find the best move. The classical evaluation computes this value as a function of various chess concepts, handcrafted by experts, tested and tuned using fishtest. The NNUE evaluation computes this value with a neural network based on basic inputs. The network is optimized and trained on the evalutions of millions of positions at moderate search depth. The NNUE evaluation was first introduced in shogi, and ported to Stockfish afterward. It can be evaluated efficiently on CPUs, and exploits the fact that only parts of the neural network need to be updated after a typical chess move. [The nodchip repository](https://github.com/nodchip/Stockfish) provides additional tools to train and develop the NNUE networks. This patch is the result of contributions of various authors, from various communities, including: nodchip, ynasu87, yaneurao (initial port and NNUE authors), domschl, FireFather, rqs, xXH4CKST3RXx, tttak, zz4032, joergoster, mstembera, nguyenpham, erbsenzaehler, dorzechowski, and vondele. This new evaluation needed various changes to fishtest and the corresponding infrastructure, for which tomtor, ppigazzini, noobpwnftw, daylen, and vondele are gratefully acknowledged. The first networks have been provided by gekkehenker and sergiovieri, with the latter net (nn-97f742aaefcd.nnue) being the current default. The evaluation function can be selected at run time with the `Use NNUE` (true/false) UCI option, provided the `EvalFile` option points the the network file (depending on the GUI, with full path). The performance of the NNUE evaluation relative to the classical evaluation depends somewhat on the hardware, and is expected to improve quickly, but is currently on > 80 Elo on fishtest: 60000 @ 10+0.1 th 1 https://tests.stockfishchess.org/tests/view/5f28fe6ea5abc164f05e4c4c ELO: 92.77 +-2.1 (95%) LOS: 100.0% Total: 60000 W: 24193 L: 8543 D: 27264 Ptnml(0-2): 609, 3850, 9708, 10948, 4885 40000 @ 20+0.2 th 8 https://tests.stockfishchess.org/tests/view/5f290229a5abc164f05e4c58 ELO: 89.47 +-2.0 (95%) LOS: 100.0% Total: 40000 W: 12756 L: 2677 D: 24567 Ptnml(0-2): 74, 1583, 8550, 7776, 2017 At the same time, the impact on the classical evaluation remains minimal, causing no significant regression: sprt @ 10+0.1 th 1 https://tests.stockfishchess.org/tests/view/5f2906a2a5abc164f05e4c5b LLR: 2.94 (-2.94,2.94) {-6.00,-4.00} Total: 34936 W: 6502 L: 6825 D: 21609 Ptnml(0-2): 571, 4082, 8434, 3861, 520 sprt @ 60+0.6 th 1 https://tests.stockfishchess.org/tests/view/5f2906cfa5abc164f05e4c5d LLR: 2.93 (-2.94,2.94) {-6.00,-4.00} Total: 10088 W: 1232 L: 1265 D: 7591 Ptnml(0-2): 49, 914, 3170, 843, 68 The needed networks can be found at https://tests.stockfishchess.org/nns It is recommended to use the default one as indicated by the `EvalFile` UCI option. Guidelines for testing new nets can be found at https://github.com/glinscott/fishtest/wiki/Creating-my-first-test#nnue-net-tests Integration has been discussed in various issues: https://github.com/official-stockfish/Stockfish/issues/2823 https://github.com/official-stockfish/Stockfish/issues/2728 The integration branch will be closed after the merge: https://github.com/official-stockfish/Stockfish/pull/2825 https://github.com/official-stockfish/Stockfish/tree/nnue-player-wip closes https://github.com/official-stockfish/Stockfish/pull/2912 This will be an exciting time for computer chess, looking forward to seeing the evolution of this approach. Bench: 4746616
Copyright 2011–2024 Next Chess Move LLC
Feedback & Support Terms of Service Privacy Policy