Dev Builds » 20180101-1217

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 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 06:01:53 2001670 2453 945 80 1428 +128.01 ± 8.51
ncm-et-4 06:01:24 2000202 2446 939 72 1435 +128.74 ± 8.45
ncm-et-5 01:53:11 1988898 767 284 24 459 +122.62 ± 14.89
ncm-et-6 01:53:07 2026145 782 317 25 440 +136.32 ± 15.46
ncm-et-7 01:52:53 2000216 774 323 21 430 +143.15 ± 15.62
ncm-et-8 01:53:12 2016530 769 279 35 455 +114.18 ± 15.16
ncm-et-9 06:01:45 2001479 2472 968 76 1428 +131.28 ± 8.51
ncm-et-10 06:02:12 1984607 2424 929 84 1411 +126.41 ± 8.57
ncm-et-11 01:53:07 2002447 766 292 14 460 +132.11 ± 14.68
ncm-et-12 01:53:05 2007779 768 305 20 443 +135.39 ± 15.24
ncm-et-13 06:01:55 1979083 2399 953 80 1366 +132.5 ± 8.75
ncm-et-14 01:52:56 2018885 759 294 30 435 +126.11 ± 15.57
ncm-et-15 06:02:01 1976726 2422 960 67 1395 +134.43 ± 8.6
20001 7788 628 11585 +130.14 ± 2.99

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
54396 ncm-et-13 2018-09-19 17:51 00:19:23 1987667 125 52 6 67 +134.15 ± 40.64
54395 ncm-et-15 2018-09-19 17:47 00:23:22 1948290 151 62 6 83 +135.3 ± 36.12
54394 ncm-et-10 2018-09-19 17:46 00:24:14 1980129 166 57 3 106 +117.28 ± 29.98
54393 ncm-et-4 2018-09-19 17:43 00:26:33 1991617 181 79 4 98 +153.17 ± 32.88
54392 ncm-et-3 2018-09-19 17:41 00:28:38 1987981 183 69 8 106 +120.41 ± 31.63
54391 ncm-et-9 2018-09-19 17:41 00:28:50 1988770 195 75 5 115 +130.53 ± 29.8
54390 ncm-et-13 2018-09-19 16:32 01:17:30 1897518 500 205 18 277 +136.56 ± 19.62
54389 ncm-et-15 2018-09-19 16:30 01:15:45 1951403 500 209 17 274 +140.62 ± 19.74
54388 ncm-et-10 2018-09-19 16:28 01:16:27 1930907 500 198 19 283 +130.15 ± 19.34
54387 ncm-et-4 2018-09-19 16:28 01:14:34 1987667 500 199 11 290 +137.37 ± 18.75
54386 ncm-et-9 2018-09-19 16:26 01:13:26 1990823 500 200 25 275 +126.97 ± 19.89
54385 ncm-et-3 2018-09-19 16:26 01:14:16 1992406 500 194 15 291 +130.15 ± 18.83
54384 ncm-et-13 2018-09-19 15:17 01:13:50 1984683 500 202 13 285 +138.18 ± 19.07
54383 ncm-et-15 2018-09-19 15:14 01:14:51 1942579 500 188 8 304 +130.94 ± 17.94
54382 ncm-et-10 2018-09-19 15:13 01:13:41 1980285 500 197 12 291 +134.95 ± 18.73
54381 ncm-et-4 2018-09-19 15:13 01:13:39 1990982 500 175 14 311 +116.0 ± 17.82
54380 ncm-et-9 2018-09-19 15:12 01:13:33 1989876 500 205 14 281 +139.81 ± 19.3
54379 ncm-et-3 2018-09-19 15:12 01:13:08 1990032 500 202 14 284 +137.37 ± 19.15
54378 ncm-et-13 2018-09-19 13:58 01:18:05 1958718 500 193 18 289 +126.97 ± 19.02
54377 ncm-et-3 2018-09-19 13:58 01:12:55 1988296 500 186 18 296 +121.46 ± 18.67
54376 ncm-et-4 2018-09-19 13:58 01:14:09 1990032 500 206 16 278 +138.99 ± 19.51
54375 ncm-et-15 2018-09-19 13:58 01:15:00 1979506 500 212 18 270 +142.26 ± 19.98
54374 ncm-et-10 2018-09-19 13:58 01:14:37 1979045 500 194 18 288 +127.76 ± 19.07
54373 ncm-et-9 2018-09-19 13:58 01:13:03 1989244 500 205 11 284 +142.26 ± 19.05
7840 ncm-et-10 2018-03-13 01:29 00:38:17 2019858 258 98 11 149 +121.93 ± 26.65
7839 ncm-et-14 2018-03-13 01:28 00:38:53 2020999 259 101 11 147 +125.98 ± 26.94
7838 ncm-et-4 2018-03-13 01:28 00:38:43 2024590 265 87 9 169 +105.38 ± 24.09
7837 ncm-et-8 2018-03-13 01:27 00:39:33 2023445 269 98 12 159 +115.11 ± 25.67
7836 ncm-et-15 2018-03-13 01:27 00:39:49 2019533 271 99 6 166 +124.27 ± 24.4
7835 ncm-et-11 2018-03-13 01:27 00:40:01 2018070 266 106 4 156 +140.4 ± 25.33
7834 ncm-et-6 2018-03-13 01:27 00:40:04 2026064 282 107 11 164 +123.19 ± 25.29
7833 ncm-et-7 2018-03-13 01:27 00:40:17 2025408 274 115 7 152 +144.78 ± 26.3
7832 ncm-et-9 2018-03-13 01:27 00:40:23 2024590 277 100 9 168 +118.53 ± 24.56
7831 ncm-et-13 2018-03-13 01:26 00:40:34 2024100 274 109 11 154 +130.01 ± 26.34
7830 ncm-et-12 2018-03-13 01:26 00:40:32 1992765 268 103 6 159 +131.72 ± 25.2
7829 ncm-et-3 2018-03-13 01:26 00:40:27 2025408 270 115 8 147 +145.66 ± 26.94
7828 ncm-et-5 2018-03-13 01:26 00:40:56 1952553 267 97 13 157 +113.14 ± 25.94
7827 ncm-et-4 2018-03-13 00:13 01:13:46 2016328 500 193 18 289 +126.97 ± 19.02
7826 ncm-et-9 2018-03-13 00:13 01:12:30 2025572 500 183 12 305 +123.81 ± 18.04
7825 ncm-et-7 2018-03-13 00:13 01:12:36 1975025 500 208 14 278 +142.26 ± 19.45
7824 ncm-et-14 2018-03-13 00:12 01:14:03 2016771 500 193 19 288 +126.18 ± 19.1
7823 ncm-et-13 2018-03-13 00:12 01:12:33 2021814 500 192 14 294 +129.35 ± 18.65
7822 ncm-et-10 2018-03-13 00:12 01:14:56 2017420 500 185 21 294 +118.33 ± 18.85
7821 ncm-et-11 2018-03-13 00:12 01:13:06 1986825 500 186 10 304 +127.76 ± 18.02
7820 ncm-et-3 2018-03-13 00:12 01:12:29 2025899 500 179 17 304 +116.78 ± 18.25
7819 ncm-et-15 2018-03-13 00:12 01:13:14 2019046 500 190 12 298 +129.35 ± 18.39
7818 ncm-et-8 2018-03-13 00:12 01:13:39 2009616 500 181 23 296 +113.68 ± 18.8
7817 ncm-et-12 2018-03-13 00:12 01:12:33 2022793 500 202 14 284 +137.37 ± 19.15
7816 ncm-et-5 2018-03-13 00:12 01:12:15 2025244 500 187 11 302 +127.76 ± 18.15
7815 ncm-et-6 2018-03-13 00:12 01:13:03 2026226 500 210 14 276 +143.89 ± 19.55

Commit

Commit ID 2ba47416cbdd5db2c7c79257072cd8675b61721f
Author Joost VandeVondele
Date 2018-01-01 12:17:51 UTC
Explicitly zero TT upon resize. as discussed in issue #1349, the way pages are allocated with calloc might imply some overhead on first write. This overhead can be large and slow down the first search after a TT resize significantly, especially for large TT. Using an explicit clear of the TT on resize fixes this problem. Not implemented, but possibly useful for large TT, is to do this zero-ing using all search threads. Not only would this be faster, it could also lead to a more favorable memory allocation on numa systems with a first touch policy. No functional change.
Copyright 2011–2024 Next Chess Move LLC