Stockfish Development Versions are build automatically if there are changes on the master branch in the git repository (https://github.com/official-stockfish/Stockfish). Use it at your own risk.
They are compiled with gcc 11.2/mingw 10 on Ubuntu 22.04.

Windows x64 for Haswell CPUs
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: CSTENTOR
Date: Sat Aug 2 16:12:12 2025 +0200
Timestamp: 1754143932

Make Resetting the Aspiration Window after FailLow more aggressive

Sets beta=(3alpha+beta)/4 when failLow. Before it was beta=(alpha+beta)/2, even
though in theory we should be getting an upper bound from a failLow, we can’t
trust the score that caused the failLow. Therefore beta=alpha doesn’t work. I
made the buffer between the new beta to the bestValue that caused the failLow
smaller.

passed STC:
https://tests.stockfishchess.org/tests/view/688674947b562f5f7b7315b5
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 119616 W: 31253 L: 30818 D: 57545 Elo +1.26
Ptnml(0-2): 313, 14028, 30724, 14397, 346

passed LTC:
https://tests.stockfishchess.org/tests/view/6887e0ad7b562f5f7b731afc
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 96978 W: 24905 L: 24466 D: 47607 Elo +1.57
Ptnml(0-2): 59, 10376, 27183, 10809, 62

closes https://github.com/official-stockfish/Stockfish/pull/6197

bench: 3085519
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: aronpetkovski
Date: Sat Aug 2 16:02:41 2025 +0200
Timestamp: 1754143361

Only do IIR in minimally reduced nodes

Passed STC
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 164256 W: 42799 L: 42299 D: 79158 Elo +1.06
Ptnml(0-2): 451, 19305, 42087, 19863, 422
https://tests.stockfishchess.org/tests/view/68868e9b7b562f5f7b731615

Passed LTC
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 279396 W: 71871 L: 71062 D: 136463 Elo +1.01
Ptnml(0-2): 126, 30117, 78404, 30924, 127
https://tests.stockfishchess.org/tests/view/6886a4977b562f5f7b731663

closes https://github.com/official-stockfish/Stockfish/pull/6196

Bench: 3490609
see source
Windows x64 for Haswell CPUs
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: J. Oster
Date: Sat Aug 2 15:54:10 2025 +0200
Timestamp: 1754142850

Display upper/lowerbound before wdl

Bound info belongs directly to score info

closes https://github.com/official-stockfish/Stockfish/pull/6195

No functional change.
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers
Windows x64
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: MinetaS
Date: Sat Aug 2 15:52:24 2025 +0200
Timestamp: 1754142744

Increase CI timeouts for perft

might fix CI failures on macOS instances

closes https://github.com/official-stockfish/Stockfish/pull/6194

No functional change
see source
Windows x64 for modern computers + AVX2
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: FauziAkram
Date: Sat Aug 2 15:50:17 2025 +0200
Timestamp: 1754142617

Re-adding ttHit condition, but this time in the inner block

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 150240 W: 39305 L: 38819 D: 72116 Elo +1.12
Ptnml(0-2): 536, 17541, 38494, 17999, 550
https://tests.stockfishchess.org/tests/view/6887ce577b562f5f7b731a85

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 159120 W: 41025 L: 40461 D: 77634 Elo +1.23
Ptnml(0-2): 96, 16974, 44866, 17518, 106
https://tests.stockfishchess.org/tests/view/6887f0ba7b562f5f7b731c5d

closes https://github.com/official-stockfish/Stockfish/pull/6193

bench: 3498926
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
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: Sat Aug 2 15:49:35 2025 +0200
Timestamp: 1754142575

Optimize find_nnz() using VBMI2

about a 0.7% speedup for the x86-64-avx512icl ARCH

closes https://github.com/official-stockfish/Stockfish/pull/6186

No functional change
see source
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: aronpetkovski
Date: Mon Jul 28 20:18:55 2025 +0200
Timestamp: 1753726735

Simplify LMR extension limit formula

This patch simplifies the maximum depth formula that reductions in LMR can allow.

Passed STC
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 35616 W: 9358 L: 9139 D: 17119 Elo +2.14
Ptnml(0-2): 101, 4051, 9278, 4284, 94
https://tests.stockfishchess.org/tests/view/688608cf966ed85face24882

Passed LTC
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 22284 W: 5833 L: 5613 D: 10838 Elo +3.43
Ptnml(0-2): 8, 2367, 6173, 2585, 9
https://tests.stockfishchess.org/tests/view/68860d7f966ed85face248d5

closes https://github.com/official-stockfish/Stockfish/pull/6191

Bench: 2902492
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Daniel Monroe
Date: Mon Jul 28 20:13:15 2025 +0200
Timestamp: 1753726395

Simplify extra continuation history updates

Passed simplification STC
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 254464 W: 65774 L: 65793 D: 122897 Elo -0.03
Ptnml(0-2): 699, 30087, 65638, 30150, 658
https://tests.stockfishchess.org/tests/view/68863283966ed85face24a59

Passed simplification LTC
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 144750 W: 37111 L: 37018 D: 70621 Elo +0.22
Ptnml(0-2): 79, 15676, 40764, 15785, 71
https://tests.stockfishchess.org/tests/view/6886655f7b562f5f7b731359

closes https://github.com/official-stockfish/Stockfish/pull/6189

Bench: 3071078
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Styx
Date: Mon Jul 28 20:11:40 2025 +0200
Timestamp: 1753726300

Simplify NMP condition

As this is tried only for cutNode, and all children of cutNodes will be nonCutNodes, the guard condition is redundant.
Simplification just removes the unnecessary condition (always true).

closes https://github.com/official-stockfish/Stockfish/pull/6187

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: 87
Date: Mon Jul 28 20:05:26 2025 +0200
Timestamp: 1753725926

Remove unnecessary deque allocation in perft

closes https://github.com/official-stockfish/Stockfish/pull/6182

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: FauziAkram
Date: Mon Jul 28 20:03:05 2025 +0200
Timestamp: 1753725785

Simplify newDepth modification formulas

Passed STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 21856 W: 5709 L: 5471 D: 10676 Elo +3.78
Ptnml(0-2): 77, 2466, 5589, 2734, 62
https://tests.stockfishchess.org/tests/view/68862ea4966ed85face24a27

Passed LTC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 108828 W: 27893 L: 27763 D: 53172 Elo +0.42
Ptnml(0-2): 59, 11860, 30435, 12012, 48
https://tests.stockfishchess.org/tests/view/68863046966ed85face24a3b

closes https://github.com/official-stockfish/Stockfish/pull/6181

Bench: 3069051
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: Nonlinear2
Date: Mon Jul 28 19:52:40 2025 +0200
Timestamp: 1753725160

increase futility value when capturing last moved piece

Passed STC:
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 204320 W: 53059 L: 52500 D: 98761 Elo +0.95
Ptnml(0-2): 551, 23874, 52778, 24379, 578
https://tests.stockfishchess.org/tests/view/688618ae966ed85face24976

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 138582 W: 35748 L: 35225 D: 67609 Elo +1.31
Ptnml(0-2): 62, 14874, 38900, 15389, 66
https://tests.stockfishchess.org/tests/view/68862717966ed85face249f2

closes https://github.com/official-stockfish/Stockfish/pull/6190

Bench: 2894413
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: FauziAkram
Date: Mon Jul 28 19:49:05 2025 +0200
Timestamp: 1753724945

Add depth condition

Passed STC:
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 103360 W: 26941 L: 26530 D: 49889 Elo +1.38
Ptnml(0-2): 291, 11964, 26770, 12353, 302
https://tests.stockfishchess.org/tests/view/68863547966ed85face24a6f

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 115128 W: 29734 L: 29258 D: 56136 Elo +1.44
Ptnml(0-2): 48, 12406, 32182, 12878, 50
https://tests.stockfishchess.org/tests/view/68864f867b562f5f7b7312f2

closes https://github.com/official-stockfish/Stockfish/pull/6184

Bench: 2972498
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
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: FauziAkram
Date: Mon Jul 28 19:45:13 2025 +0200
Timestamp: 1753724713

Tweak the logic for setting the improving flag

Tweak the logic for setting the improving flag when the static evaluation is significantly higher than alpha.

Passed STC:
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 92320 W: 24057 L: 23664 D: 44599 Elo +1.48
Ptnml(0-2): 247, 10689, 23893, 11086, 245
https://tests.stockfishchess.org/tests/view/68860aba966ed85face248a1

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 281292 W: 72496 L: 71683 D: 137113 Elo +1.00
Ptnml(0-2): 135, 30289, 78995, 31082, 145
https://tests.stockfishchess.org/tests/view/6886168c966ed85face24962

closes https://github.com/official-stockfish/Stockfish/pull/6180

Bench: 3200439
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
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 Jul 28 19:40:15 2025 +0200
Timestamp: 1753724415

Simplify static evaluation difference to move ordering logic

Passed STC simplification bounds
https://tests.stockfishchess.org/tests/view/686fcb091451bd6fb83082bc
LLR: 2.92 (-2.94,2.94) <-1.75,0.25>
Total: 762816 W: 196717 L: 197295 D: 368804 Elo -0.26
Ptnml(0-2): 2052, 90457, 196853, 90109, 1937

Passed LTC simplification bounds
https://tests.stockfishchess.org/tests/view/68808d316e17e7fa3939b35d
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 44736 W: 11499 L: 11303 D: 21934 Elo +1.52
Ptnml(0-2): 24, 4766, 12586, 4974, 18

closes https://github.com/official-stockfish/Stockfish/pull/6176

bench: 2684407
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
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: Guenther Demetz
Date: Mon Jul 28 19:35:28 2025 +0200
Timestamp: 1753724128

Simplify search functions according DRY principle

Passed STC non-regression bounds
https://tests.stockfishchess.org/tests/view/6870db4bfa93cf16d3bb286a
LLR: 3.00 (-2.94,2.94) <-1.75,0.25>
Total: 182016 W: 46735 L: 46673 D: 88608 Elo +0.12
Ptnml(0-2): 368, 19895, 50465, 19867, 413

closes https://github.com/official-stockfish/Stockfish/pull/6161

no functional change
see source
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Muzhen Gaming
Date: Mon Jul 28 19:33:00 2025 +0200
Timestamp: 1753723980

VVLTC Search Tune

alues were tuned with 125k VVLTC games.

Passed VVLTC 1st sprt:
https://tests.stockfishchess.org/tests/view/68865feb7b562f5f7b731341
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 39640 W: 10380 L: 10109 D: 19151 Elo +2.38
Ptnml(0-2): 5, 3556, 12424, 3833, 2

Passed VVLTC 2nd sprt:
https://tests.stockfishchess.org/tests/view/68864dfa7b562f5f7b7312ee
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 25972 W: 6807 L: 6537 D: 12628 Elo +3.61
Ptnml(0-2): 0, 2294, 8132, 2556, 4

closes https://github.com/official-stockfish/Stockfish/pull/6188

Bench: 3143696
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: aronpetkovski
Date: Thu Jul 24 10:38:13 2025 +0200
Timestamp: 1753346293

Simplify eval >= beta condition from NMP

It seems that now only checking if static eval is above beta by some margin is all that's necessary.

Passed Non-Regression STC
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 227264 W: 58430 L: 58421 D: 110413 Elo +0.01
Ptnml(0-2): 532, 26559, 59454, 26542, 545
https://tests.stockfishchess.org/tests/view/68763a77432ca24f6388c766

Passed Non-Regression LTC
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 72048 W: 18622 L: 18455 D: 34971 Elo +0.81
Ptnml(0-2): 26, 7775, 20272, 7908, 43
https://tests.stockfishchess.org/tests/view/687c9f2b6e17e7fa3939b0c5

closes https://github.com/official-stockfish/Stockfish/pull/6175

Bench: 2759840
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
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: Daniel Monroe
Date: Thu Jul 24 10:32:41 2025 +0200
Timestamp: 1753345961

Simplify away a term in see pruning

This change only sets the margin to 0 when depth <= 2 and margin <= 42 which
never has a difference, thus non-functional

Passed non-regression STC
LLR: 3.51 (-2.94,2.94) <-1.75,0.25>
Total: 306976 W: 78935 L: 78961 D: 149080 Elo -0.03
Ptnml(0-2): 630, 34097, 84067, 34057, 637
https://tests.stockfishchess.org/tests/view/68708015fa93cf16d3bb2782

closes https://github.com/official-stockfish/Stockfish/pull/6170

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: Thu Jul 24 10:28:17 2025 +0200
Timestamp: 1753345697

Prevent accidential misuse of TUNE()

Writing a TUNE expression like

int smallNetThreshold = 962;
TUNE(smallNetThreshold, 850, 1050);

is wrong and can lead to a weird binary. It should thus fail to even compile,
with the proper intellisense you'll also see the error directly in your editor.

A cheap way to prevent this is to disallow any fundamental type in the parameter list.

closes https://github.com/official-stockfish/Stockfish/pull/6164

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: FauziAkram
Date: Thu Jul 24 10:26:15 2025 +0200
Timestamp: 1753345575

Correct comment

closes https://github.com/official-stockfish/Stockfish/pull/6159

No functional change
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 for modern computers
Windows x64 + SSSE3
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Daniel Samek
Date: Thu Jul 24 10:24:08 2025 +0200
Timestamp: 1753345448

Add offsets to history updates.

All parameters were tuned on 60k STC games (https://tests.stockfishchess.org/tests/view/6867ef49fe0f2fe354c0c735).

Passed STC:
LLR: 2.98 (-2.94,2.94) <0.00,2.00>
Total: 60576 W: 15794 L: 15444 D: 29338 Elo +2.01
Ptnml(0-2): 141, 7003, 15642, 7369, 133
https://tests.stockfishchess.org/tests/view/686d5af11451bd6fb830772a

Passed LTC:
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 58722 W: 15159 L: 14799 D: 28764 Elo +2.13
Ptnml(0-2): 19, 6259, 16455, 6599, 29
https://tests.stockfishchess.org/tests/view/686f9fd51451bd6fb83081c9

closes https://github.com/official-stockfish/Stockfish/pull/6158

bench: 2706299
see source
Windows x64 for Haswell CPUs
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: FauziAkram
Date: Thu Jul 24 10:19:10 2025 +0200
Timestamp: 1753345150

Remove !ttData.move condition from cutNode reduction

Passed STC:
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 118080 W: 30427 L: 30298 D: 57355 Elo +0.38
Ptnml(0-2): 290, 13880, 30561, 14029, 280
https://tests.stockfishchess.org/tests/view/6853f9bc038630d25f468670

Passed LTC:
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 232272 W: 59187 L: 59182 D: 113903 Elo +0.01
Ptnml(0-2): 111, 25430, 65014, 25505, 76
https://tests.stockfishchess.org/tests/view/68585c40a596a06817bb922e

closes https://github.com/official-stockfish/Stockfish/pull/6157

bench: 2227361
see source
Windows x64 for Haswell CPUs
Windows x64 for modern computers + AVX2
Windows x64 + SSSE3
Linux x64 for Haswell CPUs
Linux x64 for modern computers + AVX2
Linux x64 for modern computers
Linux x64 + SSSE3
Linux x64
Author: Shawn Xu
Date: Thu Jul 24 10:13:24 2025 +0200
Timestamp: 1753344804

Consistent syntax for class members

Currently, search.cpp uses three different syntax when accessing class member variables

* thisThread object references. This was a remnant from when pos.this_thread() was necessary for thread-local variable access
* this object references
* implicit member variable access

This PR aims to deprecate thisThread and standardize this syntax to implicit variable access, which is consistent with the rest of SF's codebase.

closes https://github.com/official-stockfish/Stockfish/pull/6154

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: 87
Date: Thu Jul 24 10:08:35 2025 +0200
Timestamp: 1753344515

Speedup movegen with VBMI2

Passed STC
LLR: 2.96 (-2.94,2.94) <0.00,2.00>
Total: 166720 W: 43191 L: 42701 D: 80828 Elo +1.02
Ptnml(0-2): 348, 18567, 45069, 18999, 377
https://tests.stockfishchess.org/tests/view/686ae98dfe0f2fe354c0c867

Refactor movegen to emit to a vector with 16-bit elements, which enables a
speedup with AVX512-VBMI2 when writing moves to the move list.

Very crude timing measurements of perft via timing ./stockfish "go perft 7"
demonstrates approximately 17% perft speedup:

Summary
./Stockfish-dev/src/stockfish 'go perft 7' ran
1.17 ± 0.04 times faster than ./Stockfish-base/src/stockfish 'go perft 7'

Estimated overall nps increase of 0.4% via speedtest: 33605229 -> 33749825
(many thanks JonathanHallstrom).

The corresponding arch is avx512icl as it is a good baseline for consumer
avx-512 feature set support; Intel Ice Lake was the first consumer AVX-512 CPU
and it is a decent subset of what AMD Zen 4 supports.

closes https://github.com/official-stockfish/Stockfish/pull/6153

No functional change
see source

< prev page next page >