Dev Builds » 20151103-1027

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 05:53:31 1986013 2442 261 339 1842 -11.1 +/- 6.82
ncm-et-4 05:53:50 1969912 2412 258 359 1795 -14.56 +/- 7.0
ncm-et-5 01:50:12 2019706 779 61 109 609 -21.44 +/- 11.35
ncm-et-6 01:50:05 2026718 775 72 119 584 -21.1 +/- 12.1
ncm-et-7 01:49:57 2003110 770 82 107 581 -11.28 +/- 12.15
ncm-et-8 01:50:04 1972531 741 76 108 557 -15.01 +/- 12.45
ncm-et-9 05:53:31 1990431 2465 227 357 1881 -18.34 +/- 6.66
ncm-et-10 05:53:37 1985682 2437 267 335 1835 -9.7 +/- 6.85
ncm-et-11 01:50:14 1996424 757 100 105 552 -2.29 +/- 12.89
ncm-et-12 01:50:09 2020673 763 89 106 568 -7.74 +/- 12.46
ncm-et-13 05:53:24 1989092 2444 261 358 1825 -13.8 +/- 6.92
ncm-et-14 01:50:15 2020265 778 83 108 587 -11.17 +/- 12.09
ncm-et-15 05:53:40 1991376 2437 256 337 1844 -11.55 +/- 6.8
20000 2093 2847 15060 -13.1 +/- 2.39

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
72470 ncm-et-4 2019-03-04 14:51 00:21:15 1917419 142 15 23 104 -19.59 +/- 29.55
72469 ncm-et-10 2019-03-04 14:48 00:23:47 1972558 153 18 20 115 -4.54 +/- 27.49
72468 ncm-et-15 2019-03-04 14:47 00:24:36 1990103 164 21 31 112 -21.21 +/- 29.95
72467 ncm-et-3 2019-03-04 14:47 00:24:47 1965072 172 15 18 139 -6.06 +/- 22.76
72466 ncm-et-9 2019-03-04 14:44 00:27:28 1971257 178 20 30 128 -19.54 +/- 27.04
72465 ncm-et-13 2019-03-04 14:44 00:27:58 1974680 191 12 20 159 -14.56 +/- 20.12
72464 ncm-et-4 2019-03-04 13:36 01:13:22 1965999 500 46 91 363 -31.35 +/- 15.84
72463 ncm-et-15 2019-03-04 13:34 01:12:22 1976850 500 50 64 386 -9.73 +/- 14.54
72462 ncm-et-3 2019-03-04 13:34 01:12:06 1964147 500 50 68 382 -12.51 +/- 14.78
72461 ncm-et-10 2019-03-04 13:34 01:13:33 1965392 500 50 77 373 -18.78 +/- 15.31
72460 ncm-et-9 2019-03-04 13:32 01:10:59 1988394 500 41 66 393 -17.39 +/- 14.05
72459 ncm-et-13 2019-03-04 13:32 01:10:49 1967849 500 50 84 366 -23.66 +/- 15.71
72458 ncm-et-4 2019-03-04 12:23 01:12:43 1973190 500 62 75 363 -9.04 +/- 15.94
72457 ncm-et-10 2019-03-04 12:20 01:12:54 1973677 500 60 68 372 -5.56 +/- 15.41
72456 ncm-et-15 2019-03-04 12:19 01:13:25 1969556 500 62 74 364 -8.34 +/- 15.89
72455 ncm-et-13 2019-03-04 12:19 01:11:41 1964301 500 56 63 381 -4.86 +/- 14.86
72454 ncm-et-9 2019-03-04 12:19 01:12:33 1961999 500 42 76 382 -23.66 +/- 14.73
72453 ncm-et-3 2019-03-04 12:18 01:14:31 1963533 500 60 68 372 -5.56 +/- 15.41
72452 ncm-et-4 2019-03-04 11:05 01:16:19 1948635 500 47 70 383 -15.99 +/- 14.7
72451 ncm-et-3 2019-03-04 11:05 01:11:55 1969730 500 50 80 370 -20.87 +/- 15.49
72450 ncm-et-9 2019-03-04 11:05 01:12:24 1974049 500 48 78 374 -20.87 +/- 15.24
72449 ncm-et-10 2019-03-04 11:05 01:13:29 1963396 500 54 77 369 -15.99 +/- 15.57
72448 ncm-et-13 2019-03-04 11:05 01:13:05 1977727 500 58 68 374 -6.95 +/- 15.29
72447 ncm-et-15 2019-03-04 11:05 01:13:10 1972194 500 43 69 388 -18.08 +/- 14.37
7475 ncm-et-8 2018-03-11 22:23 00:35:31 1990206 241 21 27 193 -8.65 +/- 19.58
7474 ncm-et-11 2018-03-11 22:21 00:37:23 2013209 257 33 30 194 +4.06 +/- 21.05
7473 ncm-et-13 2018-03-11 22:21 00:37:26 2025081 253 30 42 181 -16.49 +/- 22.83
7472 ncm-et-12 2018-03-11 22:20 00:38:07 2020673 263 32 32 199 -0.0 +/- 20.74
7471 ncm-et-4 2018-03-11 22:20 00:38:12 2023937 270 25 39 206 -18.03 +/- 20.14
7470 ncm-et-14 2018-03-11 22:20 00:38:30 2020184 278 29 33 216 -5.0 +/- 19.3
7469 ncm-et-3 2018-03-11 22:20 00:38:37 2026390 270 26 38 206 -15.45 +/- 20.15
7468 ncm-et-15 2018-03-11 22:20 00:38:37 2020184 273 27 28 218 -1.27 +/- 18.52
7467 ncm-et-7 2018-03-11 22:19 00:39:06 1980812 270 37 34 199 +3.86 +/- 21.28
7466 ncm-et-6 2018-03-11 22:19 00:39:34 2027046 275 19 43 213 -30.4 +/- 19.33
7465 ncm-et-9 2018-03-11 22:19 00:40:05 2022793 287 29 31 227 -2.42 +/- 18.39
7464 ncm-et-5 2018-03-11 22:19 00:40:05 2022958 279 22 35 222 -16.2 +/- 18.39
7463 ncm-et-10 2018-03-11 22:17 00:40:59 2017422 284 41 36 207 +6.12 +/- 21.06
7462 ncm-et-8 2018-03-11 21:07 01:14:33 1954856 500 55 81 364 -18.08 +/- 15.86
7461 ncm-et-9 2018-03-11 21:07 01:10:02 2024099 500 47 76 377 -20.17 +/- 15.06
7460 ncm-et-7 2018-03-11 21:07 01:10:51 2025409 500 45 73 382 -19.48 +/- 14.75
7459 ncm-et-6 2018-03-11 21:07 01:10:31 2026391 500 53 76 371 -15.99 +/- 15.45
7458 ncm-et-5 2018-03-11 21:07 01:10:07 2016455 500 39 74 387 -24.36 +/- 14.4
7457 ncm-et-13 2018-03-11 21:07 01:12:25 2024917 500 55 81 364 -18.08 +/- 15.86
7456 ncm-et-15 2018-03-11 21:07 01:11:30 2019370 500 53 71 376 -12.51 +/- 15.16
7455 ncm-et-11 2018-03-11 21:07 01:12:51 1979640 500 67 75 358 -5.56 +/- 16.24
7454 ncm-et-4 2018-03-11 21:07 01:11:59 1990292 500 63 61 376 +1.39 +/- 15.18
7453 ncm-et-10 2018-03-11 21:07 01:08:55 2021651 500 44 57 399 -9.04 +/- 13.68
7452 ncm-et-14 2018-03-11 21:07 01:11:45 2020347 500 54 75 371 -14.6 +/- 15.45
7451 ncm-et-12 2018-03-11 21:07 01:12:02 2020673 500 57 74 369 -11.82 +/- 15.58
7450 ncm-et-3 2018-03-11 21:07 01:11:35 2027210 500 60 67 373 -4.86 +/- 15.36

Commit

Commit ID 9c9205860c5ab0e4f3180298e3f7082be259772c
Author Marco Costalba
Date 2015-11-03 10:27:00 UTC
Get rid of timer thread Unfortunately std::condition_variable::wait_for() is not accurate in general case and the timer thread can wake up also after tens or even hundreds of millisecs after time has elapsded. CPU load, process priorities, number of concurrent threads, even from other processes, will have effect upon it. Even official documentation says: "This function may block for longer than timeout_duration due to scheduling or resource contention delays." So retire timer and use a polling scheme based on a local thread counter that counts search() calls and a small trick to keep polling frequency constant, independently from the number of threads. Tested for no regression at very fast TC 2+0.05 th 7: LLR: 2.96 (-2.94,2.94) [-3.00,1.00] Total: 32969 W: 6720 L: 6620 D: 19629 TC 2+0.05 th 1: LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 7765 W: 1917 L: 1765 D: 4083 And at STC TC, both single thread LLR: 2.96 (-2.94,2.94) [-3.00,1.00] Total: 15587 W: 3036 L: 2905 D: 9646 And with 7 threads LLR: 2.95 (-2.94,2.94) [-3.00,1.00] Total: 8149 W: 1367 L: 1227 D: 5555 bench: 8639247
Copyright 2011–2024 Next Chess Move LLC