Commit Graph
19 Commits
Author SHA1 Message Date
mstemberaandDisservin be142337d8 Accumulator cache bugfix and cleanup
STC:
https://tests.stockfishchess.org/tests/view/663068913a05f1bf7a511dc2
LLR: 2.98 (-2.94,2.94) <-1.75,0.25>
Total: 70304 W: 18211 L: 18026 D: 34067
Ptnml(0-2): 232, 7966, 18582, 8129, 243

1) Fixes a bug introduced in
   https://github.com/official-stockfish/Stockfish/pull/5194. Only one
   psqtOnly flag was used for two perspectives which was causing
   wrong entries to be cleared and marked.
2) The finny caches should be cleared like histories and not at the
   start of every search.

closes https://github.com/official-stockfish/Stockfish/pull/5203

No functional change
2024-05-01 14:17:32 +02:00
mstemberaandDisservin a129c0695b Combine remove and add in update_accumulator_refresh_cache()
Combine remove and add in update_accumulator_refresh_cache().
Move remove before add to match other parts of the code.

STC:
https://tests.stockfishchess.org/tests/view/662d96dc6115ff6764c7f4ca
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 364032 W: 94421 L: 93624 D: 175987
Ptnml(0-2): 1261, 41983, 94811, 42620, 1341

closes https://github.com/official-stockfish/Stockfish/pull/5194

Bench: 1836777
2024-04-28 21:35:48 +02:00
mstemberaandDisservin 940a3a7383 Cache small net w/ psqtOnly support
Caching the small net in the same way as the big net allows them to
share the same code path and completely removes
update_accumulator_refresh().

STC:
https://tests.stockfishchess.org/tests/view/662bfb5ed46f72253dcfed85
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 151712 W: 39252 L: 39158 D: 73302
Ptnml(0-2): 565, 17474, 39683, 17570, 564

closes https://github.com/official-stockfish/Stockfish/pull/5194

Bench: 1836777
2024-04-28 21:30:19 +02:00
mstemberaandJoost VandeVondele 94484db6e8 Avoid permuting inputs during transform()
Avoid permuting inputs during transform() and instead do it once at load time.
Affects AVX2 and newer Intel architectures only.

https://tests.stockfishchess.org/tests/view/661306613eb00c8ccc0033c7
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 108480 W: 28319 L: 27898 D: 52263
Ptnml(0-2): 436, 12259, 28438, 12662, 445

speedups measured such as e.g.

```
Result of 100 runs
==================
base (./stockfish.master       ) =    1241128  +/- 3757
test (./stockfish.patch        ) =    1247713  +/- 3689
diff                             =      +6585  +/- 2583

speedup        = +0.0053
P(speedup > 0) =  1.0000
```

closes https://github.com/official-stockfish/Stockfish/pull/5160

No functional change
2024-04-11 22:38:38 +02:00
mstemberaandDisservin 5001d49f42 Update nnue_feature_transformer.h
Unroll update_accumulator_refresh to process two
active indices simultaneously.

The compiler might not unroll effectively because
the number of active indices isn't known at
compile time.

STC https://tests.stockfishchess.org/tests/view/65faa8850ec64f0526c4fca9
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 130464 W: 33882 L: 33431 D: 63151
Ptnml(0-2): 539, 14591, 34501, 15082, 519

closes https://github.com/official-stockfish/Stockfish/pull/5125

No functional change
2024-03-26 18:06:49 +01:00
mstemberaandJoost VandeVondele 7831131591 Only evaluate the PSQT part of the small net for large evals.
Thanks to Viren6 for suggesting to set complexity to 0.

STC https://tests.stockfishchess.org/tests/view/65d7d6709b2da0226a5a203f
LLR: 2.92 (-2.94,2.94) <0.00,2.00>
Total: 328384 W: 85316 L: 84554 D: 158514
Ptnml(0-2): 1414, 39076, 82486, 39766, 1450

LTC https://tests.stockfishchess.org/tests/view/65dce6d290f639b028a54d2e
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 165162 W: 41918 L: 41330 D: 81914
Ptnml(0-2): 102, 18332, 45124, 18922, 101

closes https://github.com/official-stockfish/Stockfish/pull/5083

bench: 1504003
2024-03-03 15:29:58 +01:00
mstemberaandDisservin 531747ee78 Improve thread voting inefficiencies
Initialize the unordered map to a reasonable
number of buckets and make the move hashes well
distributed. For more see
https://github.com/official-stockfish/Stockfish/pull/4958#issuecomment-1937351190
Also make bestThreadPV and newThreadPV references
so we don't copy entire vectors.

closes https://github.com/official-stockfish/Stockfish/pull/5048

No functional change
2024-02-11 19:55:26 +01:00
mstemberaandDisservin 9699f4f79a Fix the alignment of the transformer buffer
Fixes the issue mentioned in
https://github.com/official-stockfish/Stockfish/commit/584d9efedcde330eeb96a99215552ddfb06f52ba#r138417600.
Thanks to @cj5716 and @peregrineshahin for
spotting this!

closes https://github.com/official-stockfish/Stockfish/pull/5042

No functional change
2024-02-09 19:06:25 +01:00
mstemberaandDisservin 32e46fc47f Remove some outdated SIMD functions
Since https://github.com/official-stockfish/Stockfish/pull/4391 the x2
SIMD functions no longer serve any useful purpose.

Passed non-regression STC:
https://tests.stockfishchess.org/tests/view/659cf42579aa8af82b966d55
LLR: 2.95 (-2.94,2.94) <-1.75,0.25>
Total: 67392 W: 17222 L: 17037 D: 33133
Ptnml(0-2): 207, 7668, 17762, 7851, 208

closes https://github.com/official-stockfish/Stockfish/pull/4974

No functional change
2024-01-17 18:04:29 +01:00
mstemberaandDisservin eec361f64c Simplify bad quiets
The main difference is that instead of returning the first bad quiet as
a good one we fall through. This is actually more correct and simpler
to implement.

Non regression STC:
https://tests.stockfishchess.org/tests/view/659bbb3479aa8af82b964ec7
LLR: 2.93 (-2.94,2.94) <-1.75,0.25>
Total: 150944 W: 38399 L: 38305 D: 74240
Ptnml(0-2): 485, 18042, 38298, 18188, 459

Non regression LTC:
https://tests.stockfishchess.org/tests/view/659c6e6279aa8af82b9660eb
LLR: 2.96 (-2.94,2.94) <-1.75,0.25>
Total: 192060 W: 47871 L: 47823 D: 96366
Ptnml(0-2): 144, 21912, 51845, 22010, 119

The cutoff is now -8K instead of -7.5K.
-7.5K failed. https://tests.stockfishchess.org/tests/view/659a1f4b79aa8af82b962a0e
This was likely a false negative.

closes https://github.com/official-stockfish/Stockfish/pull/4975

Bench: 1308279
2024-01-13 19:40:53 +01:00
mstemberaandDisservin a5a76a6370 Introduce BAD_QUIET movepicker stage
Split quiets into good and bad as we do with captures. When we find
the first quiet move below a certain threshold that has been sorted we
consider all subsequent quiets bad.  Inspired by @locutus2 idea to skip
bad captures.

Passed STC:
https://tests.stockfishchess.org/tests/view/6597759f79aa8af82b95fa17
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 138688 W: 35566 L: 35096 D: 68026
Ptnml(0-2): 476, 16367, 35183, 16847, 471

Passed LTC:
https://tests.stockfishchess.org/tests/view/6598583c79aa8af82b960ad0
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 84108 W: 21468 L: 21048 D: 41592
Ptnml(0-2): 38, 9355, 22858, 9755, 48

closes https://github.com/official-stockfish/Stockfish/pull/4970

Bench: 1336907
2024-01-07 13:41:50 +01:00
mstemberaandJoost VandeVondele 8366ec48ae Scale down stat bonus
STC https://tests.stockfishchess.org/tests/view/652eff58de6d262d08d33353
LLR: 2.94 (-2.94,2.94) <0.00,2.00>
Total: 88224 W: 22618 L: 22228 D: 43378
Ptnml(0-2): 282, 10177, 22783, 10609, 261

LTC https://tests.stockfishchess.org/tests/view/652fd13bde6d262d08d3481a
LLR: 2.95 (-2.94,2.94) <0.50,2.50>
Total: 143508 W: 36674 L: 36142 D: 70692
Ptnml(0-2): 92, 15240, 40534, 15820, 68

Reduces the stat bonus by 20%. Maybe future patches can tune the actual bonus calculations for different histories.

closes https://github.com/official-stockfish/Stockfish/pull/4836

bench: 1185932
2023-10-21 10:37:27 +02:00
mstemberaandJoost VandeVondele d3d0c69dc1 Remove outdated Tile naming.
cleanup variable naming after  #4816

closes #4833

No functional change
2023-10-21 10:28:55 +02:00
mstemberaandJoost VandeVondele c17a657b04 Optimize the most common update accumalator cases w/o tiling
In the most common case where we only update a single state
it's faster to not use temporary accumulation registers and tiling.
(Also includes a couple of small cleanups.)

passed STC
https://tests.stockfishchess.org/tests/view/651918e3cff46e538ee0023b
LLR: 2.95 (-2.94,2.94) <0.00,2.00>
Total: 34944 W: 8989 L: 8687 D: 17268
Ptnml(0-2): 88, 3743, 9512, 4037, 92

A simpler version
https://tests.stockfishchess.org/tests/view/65190dfacff46e538ee00155
also passed but this version is stronger still
https://tests.stockfishchess.org/tests/view/6519b95fcff46e538ee00fa2

closes https://github.com/official-stockfish/Stockfish/pull/4816

No functional change
2023-10-08 07:42:39 +02:00
mstemberaandMarco Costalba 4a7b8180ec Remove per thread instances of Endgames. (#2056)
Similar to PSQT we only need one instance of the Endgames resource. The current per thread copies are identical and read only(after initialization) so from a design point of view it doesn't make sense to have them.

Tested for no slowdown.
http://tests.stockfishchess.org/tests/view/5c94377a0ebc5925cfff43ca
LLR: 2.95 (-2.94,2.94) [-3.00,1.00]
Total: 17320 W: 3487 L: 3359 D: 10474

No functional change.
2019-05-15 10:41:58 +02:00
mstemberaandMarco Costalba 6514500236 Less king danger if we have a knight near by to defend it. (#1987)
bench: 3653942
2019-02-03 14:16:34 +01:00
mstemberaandMarco Costalba 656aad8b0c Use a bit less code to calculate hashfull() (#1830)
* Use a bit less code to calculate hashfull(). Change post increment to preincrement as is more standard
in the rest of stockfish.  Scale result so we report 100% instead of 99.9% when completely full.

No functional change.
2018-12-23 16:10:07 +01:00
mstemberaandMarco Costalba ae5d2c38e1 Turn on random access for Syzygy files in Windows (#1840)
* This is the Windows version of
https://github.com/official-stockfish/Stockfish/pull/1829

No functional change.
2018-12-23 16:09:03 +01:00
mstemberaandMarco Costalba 378c8bdbb8 Stockfish 9
Official release version of Stockfish 9

Bench 5023629
2018-01-31 11:41:09 +01:00