| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: pb00067
Date: Tue Mar 14 08:38:02 2023 +0100 Timestamp: 1678779482 Small cleanups remove some unneeded assignments, typos, incorrect comments, add authors entry. closes https://github.com/official-stockfish/Stockfish/pull/4417 no functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: mstembera
Date: Tue Mar 14 08:29:19 2023 +0100 Timestamp: 1678778959 Remove redundant condition from capture_stage() Change a non functional promotion check to an assert. closes https://github.com/official-stockfish/Stockfish/pull/4436 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Sebastian Buchwald
Date: Tue Mar 14 08:25:14 2023 +0100 Timestamp: 1678778714 Fix Makefile for clang 16 The clang 16 release will remove the -fexperimental-new-pass-manager flag (see https://github.com/llvm/llvm-project/commit/69b2b7282e92a1b576b7bd26f3b16716a5027e8e). Thus, the commit adapts the Makefile to use this flag only for older clang versions. closes https://github.com/official-stockfish/Stockfish/pull/4437 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Alfredo Menezes
Date: Tue Mar 14 08:23:50 2023 +0100 Timestamp: 1678778630 Change mode of incbin.h Keep incbin.h with the same mode as the other source files. A mode diff might show up when working with patch files or sending the source code between devices. This patch should fix such behaviour. closes https://github.com/official-stockfish/Stockfish/pull/4442 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Michael Chaly
Date: Tue Mar 14 08:22:40 2023 +0100 Timestamp: 1678778560 Do more singular extensions This patch continues trend of last VLTC tuning - as measured by dubslow most of it gains was in lowering marging in calculation of singularBeta. This patch is a manual adjustment on top of it - it lowers multiplier of depth in calculation of singularBeta even further, from 2/3 to 1,5/2,5. Was negative at STC: https://tests.stockfishchess.org/tests/view/64089c632644b62c3393fc12 Elo: -2.49 +-1.7 (95%) LOS: 0.2% Total: 40000 W: 10601 L: 10888 D: 18511 Elo -2.49 Ptnml(0-2): 123, 4580, 10875, 4305, 117 nElo: -5.03 +-3.4 (95%) PairsRatio: 0.94 Passed 180+1.8 SPRT: https://tests.stockfishchess.org/tests/view/640096dae74a12625bcf3b33 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 160952 W: 43753 L: 43342 D: 73857 Elo +0.89 Ptnml(0-2): 25, 13984, 52039, 14411, 17 Passed 60+0.6 8 threads SPRT: https://tests.stockfishchess.org/tests/view/640dca8e65775d3b539cb7f6 LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 38824 W: 10825 L: 10554 D: 17445 Elo +2.43 Ptnml(0-2): 0, 2939, 13268, 3200, 5 closes https://github.com/official-stockfish/Stockfish/pull/4443 bench 4776866 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Dubslow
Date: Tue Mar 14 00:41:42 2023 +0100 Timestamp: 1678750902 More negative extensions on nonsingular nonpv nodes. Following up the previous gainer also in this nonsingular node section of code. Credit shared with @FauziAkram for realizing this nonsingular node stuff had some potential, and @XInTheDark for reminding us that !PvNodes better handle extensions/reductions than Pv. Passed STC: https://tests.stockfishchess.org/tests/view/640a7bb32644b62c339457c3 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 136776 W: 36598 L: 36149 D: 64029 Elo +1.14 Ptnml(0-2): 439, 14834, 37384, 15301, 430 Passed LTC: https://tests.stockfishchess.org/tests/view/640c43a02644b62c3394b23c LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 79536 W: 21363 L: 20984 D: 37189 Elo +1.66 Ptnml(0-2): 28, 7525, 24285, 7900, 30 closes https://github.com/official-stockfish/Stockfish/pull/4441 Bench: 4444953 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Michael Chaly
Date: Wed Mar 8 07:16:09 2023 +0100 Timestamp: 1678256169 Do more negative extensions This patch does negatively extend transposition table move if singular search failed and tt value is not bigger than alpha. Logic is close to what we had before recent simplification of negative extensions but uses or condition instead of and condition. Passed STC: https://tests.stockfishchess.org/tests/view/6404c8102644b62c33934607 LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 119040 W: 31841 L: 31416 D: 55783 Elo +1.24 Ptnml(0-2): 356, 13070, 32292, 13397, 405 Passed LTC: https://tests.stockfishchess.org/tests/view/6405abda2644b62c33937119 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 47216 W: 12816 L: 12496 D: 21904 Elo +2.35 Ptnml(0-2): 12, 4500, 14286, 4776, 34 closes https://github.com/official-stockfish/Stockfish/pull/4430 bench 4747020 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Joost VandeVondele
Date: Wed Mar 8 07:14:49 2023 +0100 Timestamp: 1678256089 Fix TB after capture_stage fix https://github.com/official-stockfish/Stockfish/commit/5c75c1c2fbb7bb4f0bf7c44fb855c415b788cbf7 introduced a capture_stage() function, but TB usage needs a pure capture() function. closes https://github.com/official-stockfish/Stockfish/pull/4428 No functional change. see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Maxim Masiutin
Date: Wed Mar 8 07:14:07 2023 +0100 Timestamp: 1678256047 Clarify the description of the x86-64-vnni256 and x86-64-avxvnni architectures Now it is clearly explained that "x86-64-vnni256" requires full support of AVX512-VNNI, but only 256-bit operands are used. closes https://github.com/official-stockfish/Stockfish/pull/4427 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Joost VandeVondele
Date: Sun Mar 5 16:16:16 2023 +0100 Timestamp: 1678029376 Add CITATION.cff file Make the stockfish software more easily citable, for example in academic papers. fixes https://github.com/official-stockfish/Stockfish/issues/4419 closes https://github.com/official-stockfish/Stockfish/pull/4422 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: dav1312
Date: Sun Mar 5 16:15:12 2023 +0100 Timestamp: 1678029312 Update README.md Update and simplify the readme, removing duplicated and outdated stuff and pointing to the new wiki closes https://github.com/official-stockfish/Stockfish/pull/4421 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: disservin
Date: Sun Mar 5 16:14:07 2023 +0100 Timestamp: 1678029247 Add wiki to artifacts snapshot the wiki https://github.com/official-stockfish/stockfish/wiki as part of the artifacts generated. This will allow future release to include the wiki pages as a form of documentation closes https://github.com/official-stockfish/Stockfish/pull/4420 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Michael Chaly
Date: Sun Mar 5 16:06:03 2023 +0100 Timestamp: 1678028763 Fix duplicated moves generation in movepicker in a some of cases movepicker returned some moves more than once which lead to them being searched more than once. This bug was possible because of how we use queen promotions - they are generated as a captures but are not included in position function which checks if move is a capture. Thus if any refutation (killer or countermove) was a queen promotion it was searched twice - once as a capture and one as a refutation. This patch affects various things, namely stats assignments for queen promotions and other moves if best move is queen promotion, also some heuristics in search and qsearch. With this patch every queen promotion is now considered a capture. After this patch number of found duplicated moves is 0 during normal 13 depth bench run. Passed STC: https://tests.stockfishchess.org/tests/view/63f77e01e74a12625bcd87d7 LLR: 2.95 (-2.94,2.94) <-1.75,0.25> Total: 80920 W: 21455 L: 21289 D: 38176 Elo +0.71 Ptnml(0-2): 198, 8839, 22241, 8963, 219 Passed LTC: https://tests.stockfishchess.org/tests/view/63f7e020e74a12625bcd9a76 LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 89712 W: 23674 L: 23533 D: 42505 Elo +0.55 Ptnml(0-2): 24, 8737, 27202, 8860, 33 closes https://github.com/official-stockfish/Stockfish/pull/4405 bench 4681731 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Linmiao Xu
Date: Mon Feb 27 22:07:52 2023 +0100 Timestamp: 1677532072 Update default net to nn-52471d67216a.nnue Created by retraining the master net with modifications to the previous best dataset: * Improving T80 oct+nov 2022 endgame lambda accuracy by rescoring with 12-16tb of syzygy 7p tablebases * Filtering T78 jun+jul+aug 2022 with d6pv2 search to remove positions with bestmove captures or one good move * Adding T80 sep 2022 data, rescored with 16tb of 7p tablebases, unfiltered Trained with max-epoch 900, end-lambda 0.7, and early-fen-skipping 28. ``` python3 easy_train.py \ --experiment-name leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p-lambda7-sk28 \ --training-dataset /data/leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p.binpack \ --nnue-pytorch-branch linrock/nnue-pytorch/easy-train-early-fen-skipping \ --early-fen-skipping 28 \ --start-from-engine-test-net True \ --gpus "0," \ --max_epoch 900 \ --start-lambda 1.0 \ --end-lambda 0.7 \ --gamma 0.995 \ --lr 4.375e-4 \ --tui False \ --seed $RANDOM ``` Training data was rescored and d6pv2 filtered in the same way as recent best datasets. For preparing the merged training dataset: ``` python3 interleave_binpacks.py \ leela96-eval-filt-v2.binpack \ dfrc99-eval-filt-v2.binpack \ test80-oct2022-16tb7p-eval-filt-v2-d6.binpack \ test80-nov2022-12tb7p-eval-filt-v2-d6.binpack \ T79-apr2022-12tb7p-eval-filt-v2.binpack \ T79-may2022-12tb7p-eval-filt-v2.binpack \ test78-junjulaug2022-16tb7p-eval-filt-v2-d6.binpack \ T60-nov2021-12tb7p-eval-filt-v2.binpack \ T60-dec2021-12tb7p-eval-filt-v2.binpack \ T78-sep2022-12tb7p.binpack \ test77-dec2021-16gb7p.binpack \ test80-sep2022-16tb7p.binpack \ /data/leela96-dfrc99-T80octnovT79aprmayT78junjulaugT60novdec-filt-v2-T78sep12tb7p-T77decT80sep16tb7p.binpack ``` Links for downloading the training data components can be found at: https://robotmoon.com/nnue-training-data/ Local elo at 25k nodes per move: nn-epoch839.nnue : 0.6 +/- 1.4 Passed STC: https://tests.stockfishchess.org/tests/view/63f9ab4be74a12625bcdf02e LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 84656 W: 22681 L: 22302 D: 39673 Elo +1.56 Ptnml(0-2): 271, 9343, 22734, 9696, 284 Passed LTC: https://tests.stockfishchess.org/tests/view/63fa3833e74a12625bce0c0e LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 184664 W: 49933 L: 49344 D: 85387 Elo +1.11 Ptnml(0-2): 111, 17977, 55561, 18578, 105 closes https://github.com/official-stockfish/Stockfish/pull/4416 bench: 4814343 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Dubslow
Date: Mon Feb 27 22:03:50 2023 +0100 Timestamp: 1677531830 Late counter bonus: boost underestimated moves The idea here is very intuitive: since we've just proven that the move is good, then if it previously had poor stats, boost those stats more than otherwise. Passed STC: https://tests.stockfishchess.org/tests/view/63fb504ce74a12625bce4154 LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 21128 W: 5763 L: 5481 D: 9884 Elo +4.64 Ptnml(0-2): 52, 2212, 5759, 2484, 57 Passed LTC: https://tests.stockfishchess.org/tests/view/63fb7825e74a12625bce491b LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 91904 W: 24764 L: 24359 D: 42781 Elo +1.53 Ptnml(0-2): 45, 8735, 27984, 9146, 42 closes https://github.com/official-stockfish/Stockfish/pull/4415 bench 4318808 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Sebastian Buchwald
Date: Mon Feb 27 08:29:47 2023 +0100 Timestamp: 1677482987 Unify type alias declarations The commit unifies the declaration of type aliases by replacing all typedefs with corresponding using statements. closing https://github.com/official-stockfish/Stockfish/pull/4412 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Stefan Geschwentner
Date: Mon Feb 27 08:28:45 2023 +0100 Timestamp: 1677482925 NNUE accumulator update in probcut. Call the recently added hint function for NNUE accumulator update after a failed probcut search. In this case we already searched at least some captures and tt move which, however, is not sufficient for a cutoff. So it seems we have a greater chance that the full search will also have no cutoff and hence all moves have to be searched. STC: https://tests.stockfishchess.org/tests/view/63fa74a4e74a12625bce1823 LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 70096 W: 18770 L: 18423 D: 32903 Elo +1.72 Ptnml(0-2): 191, 7342, 19654, 7651, 210 To be sure that we have no heavy interaction retest on top of #4410. Rebased STC: https://tests.stockfishchess.org/tests/view/63fb2f62e74a12625bce3b03 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 137688 W: 36790 L: 36349 D: 64549 Elo +1.11 Ptnml(0-2): 397, 14373, 38919, 14702, 453 closes https://github.com/official-stockfish/Stockfish/pull/4411 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: pb00067
Date: Mon Feb 27 08:27:26 2023 +0100 Timestamp: 1677482846 Use common_parent_position hint also at PVNodes TT hits. Credits to Stefan Geschwentner (locutus2) showing that the hint is useful on PvNodes. In contrast to his test, this version avoids to use the hint when in check. I believe checking positions aren't good candidates for the hint because: - evasion moves are rather few, so a checking pos. has much less childs than a normal position - if the king has to move the NNUE eval can't use incremental updates, so the child nodes have to do a full refresh anyway. Passed STC: https://tests.stockfishchess.org/tests/view/63f9c5b1e74a12625bcdf585 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 124472 W: 33268 L: 32846 D: 58358 Elo +1.18 Ptnml(0-2): 350, 12986, 35170, 13352, 378 closes https://github.com/official-stockfish/Stockfish/pull/4410 no functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Alfredo Menezes
Date: Mon Feb 27 08:26:29 2023 +0100 Timestamp: 1677482789 Simplify condition in step 15 Remove 'ttValue <= alpha' check for negative extension in singular search. Also apply some small code style changes. STC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 127888 W: 33766 L: 33651 D: 60471 Elo +0.31 Ptnml(0-2): 303, 14082, 35089, 14137, 333 https://tests.stockfishchess.org/tests/view/63f79528e74a12625bcd8c05 LTC: LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 89048 W: 23924 L: 23782 D: 41342 Elo +0.55 Ptnml(0-2): 27, 8635, 27065, 8763, 34 https://tests.stockfishchess.org/tests/view/63f82177e74a12625bcda6f4 LTC (retest): LLR: 2.94 (-2.94,2.94) <-1.75,0.25> Total: 196360 W: 52514 L: 52475 D: 91371 Elo +0.07 Ptnml(0-2): 103, 19066, 59780, 19151, 80 https://tests.stockfishchess.org/tests/view/63f934bfe74a12625bcdd929 closes https://github.com/official-stockfish/Stockfish/pull/4407 Bench: 5310866 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Michael Chaly
Date: Fri Feb 24 20:14:44 2023 +0100 Timestamp: 1677266084 Search tuning at very long time control This patch is a result of tuning session of approximately 100k games at 120+1.2. Biggest changes are in extensions, stat bonus and depth reduction for nodes without a tt move. Failed STC: https://tests.stockfishchess.org/tests/view/63f72c72e74a12625bcd7938 LLR: -2.94 (-2.94,2.94) <0.00,2.00> Total: 13872 W: 3535 L: 3769 D: 6568 Elo -5.86 Ptnml(0-2): 56, 1621, 3800, 1419, 40 Close to neutral at LTC: https://tests.stockfishchess.org/tests/view/63f738f5e74a12625bcd7b8a Elo: 0.80 +-1.2 (95%) LOS: 90.0% Total: 60000 W: 16213 L: 16074 D: 27713 Elo +0.80 Ptnml(0-2): 24, 5718, 18379, 5853, 26 nElo: 1.82 +-2.8 (95%) PairsRatio: 1.02 Passed 180+1.8 VLTC: https://tests.stockfishchess.org/tests/view/63f868f3e74a12625bcdb33e LLR: 2.95 (-2.94,2.94) <0.50,2.50> Total: 15864 W: 4449 L: 4202 D: 7213 Elo +5.41 Ptnml(0-2): 1, 1301, 5083, 1544, 3 Passed 60+0.6 8 threads SMP VLTC: https://tests.stockfishchess.org/tests/view/63f8a5d6e74a12625bcdbdb3 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 6288 W: 1821 L: 1604 D: 2863 Elo +11.99 Ptnml(0-2): 0, 402, 2123, 619, 0 closes https://github.com/official-stockfish/Stockfish/pull/4406 bench 4705194 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Sebastian Buchwald
Date: Fri Feb 24 20:12:53 2023 +0100 Timestamp: 1677265973 Fix typo in method name closes https://github.com/official-stockfish/Stockfish/pull/4404 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Linmiao Xu
Date: Thu Feb 23 13:27:57 2023 +0100 Timestamp: 1677155277 Reintroduce nnue pawn scaling with lower lazy thresholds Params found with the nevergrad TBPSA optimizer via nevergrad4sf modified to: * use SPRT LLR with fishtest STC elo gainer bounds [0, 2] as the objective function * increase the game batch size after each new optimal point is found The params were the optimal point after TBPSA iteration 7 and 160 nevergrad evaluations with: * initial batch size of 96 games per evaluation * batch size increase of 64 games after each iteration * a budget of 512 evaluations * TC: fixed 1.5 million nodes per move, no time limit nevergrad4sf enables optimizing stockfish params with TBPSA: https://github.com/vondele/nevergrad4sf Using pentanomial game results with smaller game batch sizes was inspired by: Use of SPRT LLR calculated from pentanomial game results as the objective function was an experiment at maximizing the information from game batches to reduce the computational cost for TBPSA to converge on good parameters. For the exact code used to find the params: https://github.com/linrock/tuning-fork Passed STC: https://tests.stockfishchess.org/tests/view/63f4ef5ee74a12625bcd114a LLR: 2.94 (-2.94,2.94) <0.00,2.00> Total: 66552 W: 17736 L: 17390 D: 31426 Elo +1.81 Ptnml(0-2): 164, 7229, 18166, 7531, 186 Passed LTC: https://tests.stockfishchess.org/tests/view/63f56028e74a12625bcd2550 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 71264 W: 19150 L: 18787 D: 33327 Elo +1.77 Ptnml(0-2): 23, 6728, 21771, 7083, 27 closes https://github.com/official-stockfish/Stockfish/pull/4401 bench 3687580 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Joost VandeVondele
Date: Thu Feb 23 13:25:35 2023 +0100 Timestamp: 1677155135 Introduce a function to compute NNUE accumulator This patch introduces `hint_common_parent_position()` to signal that potentially several child nodes will require an NNUE eval. By populating explicitly the accumulator, these subsequent evaluations can be performed more efficiently. This was based on the observation that calculating the evaluation in an excluded move position yielded a significant Elo gain, even though the evaluation itself was already available (work by pb00067). Sopel wrote the code to perform just the accumulator update. This PR is based on cleaned up code that passed STC: https://tests.stockfishchess.org/tests/view/63f62f9be74a12625bcd4aa0 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 110368 W: 29607 L: 29167 D: 51594 Elo +1.39 Ptnml(0-2): 41, 10551, 33572, 10967, 53 and in an the earlier (equivalent) version passed STC: https://tests.stockfishchess.org/tests/view/63f3c3fee74a12625bcce2a6 LLR: 2.95 (-2.94,2.94) <0.00,2.00> Total: 47552 W: 12786 L: 12467 D: 22299 Elo +2.33 Ptnml(0-2): 120, 5107, 12997, 5438, 114 passed LTC: https://tests.stockfishchess.org/tests/view/63f45cc2e74a12625bccfa63 LLR: 2.94 (-2.94,2.94) <0.50,2.50> Total: 110368 W: 29607 L: 29167 D: 51594 Elo +1.39 Ptnml(0-2): 41, 10551, 33572, 10967, 53 closes https://github.com/official-stockfish/Stockfish/pull/4402 Bench: 3726250 see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Sebastian Buchwald
Date: Thu Feb 23 13:24:37 2023 +0100 Timestamp: 1677155077 Remove unused macros closes https://github.com/official-stockfish/Stockfish/pull/4397 No functional change see source |
| Windows x64 for Haswell CPUs Windows x64 for modern computers + AVX2 Windows x64 for modern computers Windows x64 + SSSE3 Windows x64 Linux x64 for Haswell CPUs Linux x64 for modern computers + AVX2 Linux x64 for modern computers Linux x64 + SSSE3 Linux x64 | Author: Sebastian Buchwald
Date: Thu Feb 23 13:22:03 2023 +0100 Timestamp: 1677154923 Add support for ARM dot product instructions The sdot instruction computes (and accumulates) a signed dot product, which is quite handy for Stockfish's NNUE code. The instruction is optional for Armv8.2 and Armv8.3, and mandatory for Armv8.4 and above. The commit adds a new 'arm-dotprod' architecture with enabled dot product support. It also enables dot product support for the existing 'apple-silicon' architecture, which is at least Armv8.5. The following local speed test was performed on an Apple M1 with ARCH=apple-silicon. I had to remove CPU pinning from the benchmark script. However, the results were still consistent: Checking both binaries against themselves reported a speedup of +0.0000 and +0.0005, respectively. ``` Result of 100 runs ================== base (...ish.037ef3e1) = 1917997 +/- 7152 test (...fish.dotprod) = 2159682 +/- 9066 diff = +241684 +/- 2923 speedup = +0.1260 P(speedup > 0) = 1.0000 CPU: 10 x arm Hyperthreading: off ``` Fixes #4193 closes https://github.com/official-stockfish/Stockfish/pull/4400 No functional change see source |