New engine releases & news H1 2022

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

User avatar
pedrox
Posts: 1056
Joined: Fri Mar 10, 2006 6:07 am
Location: Basque Country (Spain)

Re: New engine releases & news H1 2022

Post by pedrox »

Caligula 1.0b

Many changes. Little Elo.

This version has approximately the strength of DanaSah 8.3, about 2700 CCRL.

There are many differences between DanaSah (eval classic) and Caligula. For example Caligula uses things that DanaSah does not have such as piece list or pawn hash table. The search differs in things like pruning, the way of calculating things in the evaluation is different (mobility, king safety, etc). Overall the Caligula source code is about half the size of DanaSah.

https://pedrox.wixsite.com/caligula
Modern Times
Posts: 3553
Joined: Thu Jun 07, 2012 11:02 pm

Re: New engine releases & news H1 2022

Post by Modern Times »

pedrox wrote: Tue May 03, 2022 8:43 pm Caligula 1.0b
Does it play chess960 ?
User avatar
pedrox
Posts: 1056
Joined: Fri Mar 10, 2006 6:07 am
Location: Basque Country (Spain)

Re: New engine releases & news H1 2022

Post by pedrox »

Modern Times wrote: Tue May 03, 2022 8:52 pm
pedrox wrote: Tue May 03, 2022 8:43 pm Caligula 1.0b
Does it play chess960 ?
No. DanaSah yes.
User avatar
Graham Banks
Posts: 41473
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: New engine releases & news H1 2022

Post by Graham Banks »

pedrox wrote: Tue May 03, 2022 8:43 pm Caligula 1.0b

Many changes. Little Elo.

This version has approximately the strength of DanaSah 8.3, about 2700 CCRL.

There are many differences between DanaSah (eval classic) and Caligula. For example Caligula uses things that DanaSah does not have such as piece list or pawn hash table. The search differs in things like pruning, the way of calculating things in the evaluation is different (mobility, king safety, etc). Overall the Caligula source code is about half the size of DanaSah.

https://pedrox.wixsite.com/caligula
64-bit?
gbanksnz at gmail.com
User avatar
pedrox
Posts: 1056
Joined: Fri Mar 10, 2006 6:07 am
Location: Basque Country (Spain)

Re: New engine releases & news H1 2022

Post by pedrox »

Graham Banks wrote: Wed May 04, 2022 1:14 am
pedrox wrote: Tue May 03, 2022 8:43 pm Caligula 1.0b

Many changes. Little Elo.

This version has approximately the strength of DanaSah 8.3, about 2700 CCRL.

There are many differences between DanaSah (eval classic) and Caligula. For example Caligula uses things that DanaSah does not have such as piece list or pawn hash table. The search differs in things like pruning, the way of calculating things in the evaluation is different (mobility, king safety, etc). Overall the Caligula source code is about half the size of DanaSah.

https://pedrox.wixsite.com/caligula
64-bit?
32-bit, now use only bitboards for passed pawns, not will gain speed with 64 bits.
Gabor Szots
Posts: 1364
Joined: Sat Jul 21, 2018 7:43 am
Location: Szentendre, Hungary
Full name: Gabor Szots

Re: New engine releases & news H1 2022

Post by Gabor Szots »

Gabor Szots
CCRL testing group
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: New engine releases & news H1 2022

Post by Frank Quisinsky »

Hi there,

The programmer of Stash ...
Morgan Houppin (U.S.A.)
https://gitlab.com/mhouppin/stash-bot/-/releases

Released v33 for around 1 hour!

The situation:
All what you have to do ...
You have to go on Stash site and to download the engine!
I am ready!

:-)

Thank you!

Best
Frank
User avatar
Sylwy
Posts: 4469
Joined: Fri Apr 21, 2006 4:19 pm
Location: IASI - the historical capital of MOLDOVA
Full name: SilvianR

Re: New engine releases & news H1 2022

Post by Sylwy »

Frank Quisinsky wrote: Thu May 12, 2022 4:18 pm
Morgan Houppin (U.S.A.)
France :wink:

https://www.chessprogramming.org/Morgan_Houppin
Frank Quisinsky
Posts: 6808
Joined: Wed Nov 18, 2009 7:16 pm
Location: Gutweiler, Germany
Full name: Frank Quisinsky

Re: New engine releases & news H1 2022

Post by Frank Quisinsky »

Hi,

aah ...
The wrong flag in my engine overview.
I am crazy!

Thank you, and sorry!

Best
Frank
User avatar
Guenther
Posts: 4611
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: New engine releases & news H1 2022

Post by Guenther »

FoxSEE 8.0.0
https://github.com/redsalmon91/FoxSEE/r ... tag/v8.0.0
(strangely the compilation is removed currently, while I still downloaded it five hours ago - no idea if it will be recompiled)

Marvin new net net-9a7303c.nnue
https://github.com/bmdanielsson/marvin- ... ee/main/v2
(for versions > 5.1.0)

quick test comparison from startpos on my slow hardware with changed uci option
(no need for compiling, if you are not able to do so, just point to the new net file in the uci options)

Code: Select all

position startpos
go depth 18
info string Using NNUE evaluation with net-422c365.nnue
...
info depth 8 seldepth 11 nodes 13249 time 366 nps 36000 tbhits 0 hashfull 1 score cp 26 pv e2e4 e7e5 b1c3 g8f6 g1f3 b8c6 d2d4
info depth 9 seldepth 19 nodes 35980 time 460 nps 78000 tbhits 0 hashfull 4 score cp 17 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 g8f6 d4c6 b7c6 b1d2 a8b8
info depth 10 seldepth 13 nodes 42620 time 503 nps 84000 tbhits 0 hashfull 5 score cp 17 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 g8f6 d4c6 b7c6 b1d2 a8b8
info depth 11 seldepth 12 nodes 67792 time 614 nps 110000 tbhits 0 hashfull 10 score cp 21 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 g8f6 d4c6 b7c6 e4e5 f6e4
info depth 12 seldepth 20 nodes 92089 time 718 nps 128000 tbhits 0 hashfull 12 score cp 31 pv e2e4 e7e5 g1f3 b8c6 d2d4 e5d4 f3d4 g8f6 d4c6 b7c6 e4e5 d8e7 c1f4
info depth 13 seldepth 24 nodes 262826 time 1193 nps 220000 tbhits 0 hashfull 39 score cp 32 pv e2e4 e7e6 d2d4 d7d5 b1c3 f8b4 d1g4 g7g6 e4e5 h7h5 g4g3 b8c6 g1f3
info depth 14 seldepth 30 nodes 695770 time 2344 nps 296000 tbhits 0 hashfull 115 score cp 23 pv e2e4 e7e6 d2d4 d7d5 b1c3 d5e4 c3e4 b8d7 g1f3 g8f6 e4f6 d7f6 c1e3 f8d6 c2c3 h7h6 f1e2 e8g8
info depth 15 seldepth 28 nodes 1259772 time 3837 nps 328000 tbhits 0 hashfull 191 score cp 26 pv e2e4 c7c5 g1f3 e7e6 f1e2 g8f6 b1c3 b8c6 d2d4 d7d5 e4d5 f6d5 d4c5 d5c3 d1d8 c6d8 b2c3 f8c5
info depth 16 seldepth 28 nodes 1831360 time 5473 nps 334000 tbhits 0 hashfull 269 score cp 19 pv e2e4 c7c5 g1f3 e7e6 f1e2 b8c6 e1g1 g8f6 b1c3 f8e7 d2d4 c5d4 f3d4 e8g8 c1e3 a8b8 d4c6 b7c6
info depth 17 seldepth 26 nodes 3768354 time 10722 nps 351000 tbhits 0 hashfull 524 score cp 25 pv e2e4 c7c5 g1f3 e7e6 f1e2 b8c6 e1g1 g8f6 b1c3 f8e7 d2d4 c5d4 f3d4 e8g8 c1e3 d8a5 d1d2
info depth 18 seldepth 32 nodes 5000735 time 14154 nps 353000 tbhits 0 hashfull 646 score cp 24 pv e2e4 c7c5 g1f3 e7e6 f1e2 g8f6 b1c3 b8c6 e1g1 f8e7 d2d4 c5d4 f3d4 e8g8 d4c6 d7c6 h2h3 e7c5 c1g5 d8d1 f1d1
bestmove e2e4
ucinewgame
setoption name EvalFile value net-9a7303c.nnue
isready
readyok
go depth 18
info string Using NNUE evaluation with net-9a7303c.nnue
...
info depth 8 seldepth 12 nodes 10196 time 289 nps 35000 tbhits 0 hashfull 2 score cp 25 pv e2e4 c7c5 g1f3 e7e6 b1c3 b8c6 d2d4 c5d4 f3d4
info depth 9 seldepth 14 nodes 21473 time 359 nps 59000 tbhits 0 hashfull 5 score cp 24 pv e2e4 c7c5 g1f3 e7e6 b1c3 b8c6 f1e2 d7d5
info depth 10 seldepth 17 nodes 29893 time 403 nps 74000 tbhits 0 hashfull 6 score cp 24 pv e2e4 c7c5 g1f3 e7e6 c2c4 b8c6 b1c3 f8e7 d2d4 c5d4 f3d4
info depth 11 seldepth 20 nodes 58447 time 501 nps 116000 tbhits 0 hashfull 8 score cp 22 pv e2e4 c7c5 g1f3 b8c6 c2c4 g8f6 b1c3 e7e6 f1e2 d7d5 e4e5
info depth 12 seldepth 21 nodes 118902 time 680 nps 174000 tbhits 0 hashfull 15 score cp 19 pv e2e4 c7c5 g1f3 b8c6 c2c4 g8f6 b1c3 e7e6 f1e2 d7d5 e4e5 f6d7 c4d5 e6d5
info depth 13 seldepth 24 nodes 239953 time 1016 nps 236000 tbhits 0 hashfull 35 score cp 17 pv e2e4 c7c5 g1f3 b8c6 c2c4 g8f6 b1c3 e7e6 f1e2 d7d5 e4d5 e6d5 e1g1 d5d4
info depth 13 seldepth 24 nodes 431341 time 1528 nps 282000 tbhits 0 hashfull 64 score cp 20 pv c2c4 e7e6 d2d4 d7d5 g1f3 g8f6 b1c3 c7c6 e2e3 f8e7 f1e2 e8g8 e1g1 d5c4 e2c4
info depth 14 seldepth 22 nodes 801611 time 2516 nps 318000 tbhits 0 hashfull 123 score cp 32 pv e2e4 c7c5 g1f3 b8c6 b1c3 g8f6 d2d4 c5d4 f3d4 e7e6 d4c6 b7c6 f1e2 d7d5 e4e5
info depth 15 seldepth 24 nodes 940080 time 2900 nps 324000 tbhits 0 hashfull 142 score cp 25 pv e2e4 c7c5 g1f3 b8c6 b1c3 g8f6 d2d4 c5d4 f3d4 e7e6 d4c6 b7c6 f1e2 f8b4 c1d2 e8g8 e1g1
info depth 16 seldepth 27 nodes 2255037 time 6473 nps 348000 tbhits 0 hashfull 346 score cp 16 pv e2e4 e7e6 d2d4 d7d5 e4d5 e6d5 g1f3 g8f6 f1d3 f8e7 e1g1 e8g8 c2c4 b8c6 b1c3 c8e6 c4d5 f6d5
info depth 17 seldepth 28 nodes 4215838 time 25305 nps 166000 tbhits 0 hashfull 588 score cp 23 pv e2e4 e7e6 d2d4 d7d5 e4d5 e6d5 g1f3 g8f6 f1d3 f8e7 e1g1 e8g8 c2c4 c7c6 c4d5 f6d5 f1e1 d5b4 b1c3 b4d3 d1d3
info depth 18 seldepth 30 nodes 11359562 time 44432 nps 255000 tbhits 0 hashfull 957 score cp 17 pv e2e4 e7e6 d2d4 d7d5 e4d5 e6d5 g1f3 g8f6 f1d3 f8e7 e1g1 e8g8 f1e1 b8c6 c2c3 e7d6 b2b4 f8e8 b4b5 c6a5
bestmove e2e4
https://rwbc-chess.de

trollwatch:
Talkchess nowadays is a joke - it is full of trolls/idiots/people stuck in the pleistocene > 80% of the posts fall into this category...