SF PB (Stockfish)

[massimilianogoi]

Just casually noticed because I was compiling it: looks like the code is messed up: half of it is in the incbin folder..

This is the code, reviewed and corrected:

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

SF-PB-110723_x64 "Simplify PvNode reduction" https://github.com/official-stockfish/S ... b5d6fab80a

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

Hmm, once again the AVX is faster than AVX2.

Btw, i wanted to compile the latest version "Remove classical evaluation" (120723), but there is a big regression -2 elo, now we need to wait till this 2 elo will be gained, plus extra patches with extra elo.
STC https://tests.stockfishchess.org/tests/ ... a549c52157
LTC https://tests.stockfishchess.org/tests/ ... a549c5215b
VLTC https://tests.stockfishchess.org/tests/ ... a549c52168

[Skynet]

Strange... something messy in the Makefile?

Nope, everything is fine.

Just casually noticed because I was compiling it: looks like the code is messed up: half of it is in the incbin folder..

This is the code, reviewed and corrected:

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

Yeah, i was in a hurry and i put some of the code in incbin folder..

[Orgz]

Hope it doesn't offend you skynet that i posted here. i didnt want to open another sf pb post which is unnecessary. anyway

SF-POLY 190723
Same as the father Stockfish 190723 capable of using 4 bin books + a little search change. on my device the engine is better than the father. since engines behave differently on devices u'll see how he plays on yourS, thank you enjoy
https://pixeldrain.com/u/ax2F5s4a

TR

[Skynet]

Hope it doesn't offend you skynet that i posted here. i didnt want to open another sf pb post which is unnecessary. anyway

SF-POLY 190723
Same as the father Stockfish 190723 capable of using 4 bin books + a little search change. on my device the engine is better than the father. since engines behave differently on devices u'll see how he plays on yourS, thank you enjoy
https://pixeldrain.com/u/ax2F5s4a

TR

Hello, no it doesn't offend or bother me. But, since i never do any (mine) functional changes in SF-PB code, it is an Admin/Mod decision to keep your engine here or move it somewhere else.

[Skynet]

SF-PB-280723 "Malus during move ordering for putting pieces en prise" https://github.com/official-stockfish/S ... d82b75736d

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

[massimilianogoi]

Personally, I'm skeptical about code modifications, insertion of the polyglot expansion apart: if a modification was worthing, it would definitely end in the code of Stockfish. They are doing everything to make their engine growing in strength. More than one year ago I discovered a main line on C54 after hours of testing variations with a 128 threads machine (and alot of Ducats spent), AFTER tuning the Goi book with that line (I don't remember how much time passed) the variation ended in the official Stockfish NNUE network. They sue if someone uses a portion of their code to create an engine on sale but they don't have any problem stealing lines and ruin a market. This is playing dirty in my opinion. Yes, moves don't belong to anyone, sure, but I precisely remember, I can swear that Stockfish didn't move with that precise line before I discovered it and put in on my book.

[Event "?"] [Site "?"] [Date "2023.08.02"] [Round "?"] [White "Player"] [Black "Stockfish 16"] [Result "*"] [TimeControl "60/120"] 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. O-O Nxe4 5. Re1 Nd6 6. Nxe5 Be7 7. Bf1 Nxe5 8. Rxe5 O-O 9. d4 Bf6 10. Re1 Re8 11. c3 Rxe1 12. Qxe1 *

Hope it doesn't offend you skynet that i posted here. i didnt want to open another sf pb post which is unnecessary. anyway

SF-POLY 190723
Same as the father Stockfish 190723 capable of using 4 bin books + a little search change. on my device the engine is better than the father. since engines behave differently on devices u'll see how he plays on yourS, thank you enjoy
https://pixeldrain.com/u/ax2F5s4a

TR

[Skynet]

SF-PB-160723 "Add -funroll-loops to CXXFLAGS".
Note: now AVX2 and SSE41 (popcnt) became a bit faster than AVX.

Hidden Content

This board requires you to be registered and logged-in to view hidden content.

[Skynet]

SF-PB-230823 "Reduce repetitions branches". Now the engine has a logo (+10 elo guaranteed ).

Many thanks to Marco for the help, and explanation about what and how to do!!
Note, for unknown reason AVX2 became a bit slower than AVX and SSE41_popcnt.
https://krakenfiles.com/view/VtGcH0H57C/file.html
https://pixeldrain.com/u/u2g6cM2B

[Skynet]

SF-PB-240823 "Play turbulent when defending, simpler when attacking" AVX2 is the faster, again..
https://pixeldrain.com/u/Cepd9R2b

[Skynet]

SF-PB-030923 "Introduce simple_eval() for lazy evaluations."
https://pixeldrain.com/u/C44v8qcL
https://krakenfiles.com/view/0DOc6RJD19/file.html