Dev Builds » 20200127-1016

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 07:45:56 1962679 3071 1941 57 1073 +248.24 ± 10.32
ncm-et-4 07:45:39 1966830 3114 1847 57 1210 +227.46 ± 9.67
ncm-et-9 08:59:44 1970667 3576 2262 62 1252 +249.2 ± 9.55
ncm-et-10 09:01:02 1962070 3573 2219 53 1301 +244.22 ± 9.34
ncm-et-13 07:45:07 1970377 3082 1865 53 1164 +234.34 ± 9.87
ncm-et-15 08:59:58 1969602 3585 2209 80 1296 +237.51 ± 9.4
20001 12343 362 7296 +240.29 ± 3.95

Test Detail

ID Host Started (UTC) Duration Base NPS Games Wins Losses Draws Elo CLI PGN
113509 ncm-et-3 2020-01-28 09:54 00:10:57 1962154 71 49 2 20 +276.67 ± 79.2
113508 ncm-et-10 2020-01-28 09:54 00:11:36 1966023 73 48 2 23 +257.67 ± 73.24
113507 ncm-et-9 2020-01-28 09:53 00:11:47 1972981 76 51 2 23 +266.22 ± 73.33
113506 ncm-et-13 2020-01-28 09:53 00:12:16 1967240 82 48 1 33 +226.61 ± 59.21
113505 ncm-et-15 2020-01-28 09:52 00:13:00 1968007 85 56 0 29 +274.72 ± 63.46
113504 ncm-et-4 2020-01-28 09:49 00:16:38 1963535 114 67 1 46 +229.61 ± 49.71
113503 ncm-et-9 2020-01-28 08:37 01:14:58 1968950 500 313 10 177 +244.1 ± 25.51
113502 ncm-et-3 2020-01-28 08:37 01:16:11 1962456 500 322 12 166 +251.89 ± 26.45
113501 ncm-et-10 2020-01-28 08:37 01:16:02 1966325 500 311 7 182 +245.2 ± 25.02
113500 ncm-et-13 2020-01-28 08:37 01:15:23 1973115 500 306 9 185 +237.58 ± 24.86
113499 ncm-et-15 2020-01-28 08:34 01:16:55 1973898 500 301 12 187 +229.12 ± 24.79
113498 ncm-et-4 2020-01-28 08:33 01:14:42 1968021 500 318 2 180 +258.75 ± 24.97
113497 ncm-et-3 2020-01-28 07:21 01:15:05 1964150 500 331 6 163 +269.37 ± 26.56
113496 ncm-et-9 2020-01-28 07:21 01:15:45 1973272 500 325 10 165 +257.59 ± 26.5
113495 ncm-et-13 2020-01-28 07:20 01:15:27 1969101 500 302 8 190 +234.38 ± 24.46
113494 ncm-et-10 2020-01-28 07:20 01:15:36 1961227 500 317 9 174 +249.64 ± 25.72
113493 ncm-et-15 2020-01-28 07:18 01:15:02 1968778 500 299 12 189 +227.04 ± 24.65
113492 ncm-et-4 2020-01-28 07:17 01:14:28 1968947 500 293 8 199 +224.97 ± 23.82
113491 ncm-et-3 2020-01-28 06:04 01:15:36 1964456 500 317 8 175 +250.77 ± 25.61
113490 ncm-et-13 2020-01-28 06:04 01:15:24 1971564 500 302 8 190 +234.38 ± 24.46
113489 ncm-et-10 2020-01-28 06:03 01:15:49 1965232 500 307 6 187 +241.91 ± 24.61
113488 ncm-et-15 2020-01-28 06:03 01:13:54 1971268 500 312 15 173 +237.58 ± 25.92
113487 ncm-et-9 2020-01-28 06:03 01:16:36 1965546 500 305 11 184 +234.38 ± 24.99
113486 ncm-et-4 2020-01-28 06:01 01:15:14 1969890 500 274 10 216 +204.07 ± 22.77
113485 ncm-et-3 2020-01-28 04:48 01:15:21 1963993 500 303 6 191 +237.58 ± 24.31
113484 ncm-et-13 2020-01-28 04:47 01:15:17 1972961 500 306 10 184 +236.51 ± 24.97
113483 ncm-et-10 2020-01-28 04:46 01:15:48 1960321 500 295 5 200 +230.16 ± 23.63
113482 ncm-et-15 2020-01-28 04:46 01:15:54 1961535 500 319 10 171 +250.77 ± 25.99
113481 ncm-et-9 2020-01-28 04:46 01:16:04 1965702 500 301 9 190 +232.26 ± 24.49
113480 ncm-et-4 2020-01-28 04:45 01:14:38 1962941 500 289 13 198 +215.85 ± 24.03
113479 ncm-et-13 2020-01-28 03:30 01:15:54 1965539 500 285 11 204 +213.85 ± 23.58
113478 ncm-et-4 2020-01-28 03:30 01:14:03 1970175 500 310 10 180 +240.82 ± 25.27
113477 ncm-et-3 2020-01-28 03:30 01:16:37 1957859 500 334 5 161 +274.22 ± 26.71
113476 ncm-et-15 2020-01-28 03:30 01:14:56 1969259 500 318 5 177 +255.3 ± 25.34
113475 ncm-et-10 2020-01-28 03:30 01:15:07 1959720 500 312 9 179 +244.1 ± 25.32
113474 ncm-et-9 2020-01-28 03:30 01:14:40 1972654 500 340 4 156 +282.94 ± 27.13
113428 ncm-et-9 2020-01-27 18:19 01:15:24 1972499 500 314 8 178 +247.41 ± 25.37
113427 ncm-et-15 2020-01-27 18:18 01:16:13 1969087 500 303 10 187 +233.32 ± 24.74
113426 ncm-et-10 2020-01-27 18:17 01:16:33 1957707 500 316 8 176 +249.64 ± 25.53
113425 ncm-et-4 2020-01-27 17:10 01:15:56 1964301 500 296 13 191 +222.92 ± 24.53
113424 ncm-et-13 2020-01-27 17:05 01:15:26 1973122 500 316 6 178 +251.89 ± 25.3
113423 ncm-et-9 2020-01-27 17:03 01:14:30 1973739 500 313 8 179 +246.3 ± 25.29
113422 ncm-et-15 2020-01-27 17:03 01:14:04 1974984 500 301 16 183 +224.97 ± 25.17
113421 ncm-et-3 2020-01-27 17:02 01:16:09 1963686 500 285 18 197 +206.98 ± 24.2
113420 ncm-et-10 2020-01-27 17:01 01:14:31 1960011 500 313 7 180 +247.41 ± 25.18

Commit

Commit ID 39437f4e55aaa26ef9f0d5a1c762e560e9ffde32
Author Sami Kiminki
Date 2020-01-27 10:16:10 UTC
Advise the kernel to use huge pages (Linux) Align the TT allocation by 2M to make it huge page friendly and advise the kernel to use huge pages. Benchmarks on my i7-8700K (6C/12T) box: (3 runs per bench per config) vanilla (nps) hugepages (nps) avg ================================================================================== bench | 3012490 3024364 3036331 3071052 3067544 3071052 +1.5% bench 16 12 20 | 19237932 19050166 19085315 19266346 19207025 19548758 +1.1% bench 16384 12 20 | 18182313 18371581 18336838 19381275 19738012 19620225 +7.0% On my box, huge pages have a significant perf impact when using a big hash size. They also speed up TT initialization big time: vanilla (s) huge pages (s) speed-up ======================================================================= time stockfish bench 16384 1 1 | 5.37 1.48 3.6x In practice, huge pages with auto-defrag may always be enabled in the system, in which case this patch has no effect. This depends on the values in /sys/kernel/mm/transparent_hugepage/enabled and /sys/kernel/mm/transparent_hugepage/defrag. closes https://github.com/official-stockfish/Stockfish/pull/2463 No functional change
Copyright 2011–2024 Next Chess Move LLC