Stockfish 15 is ready

[Jouni]

1 core bench is USELESS. Only for checking compile.

[Graham Banks]

Live broadcast of Gauntlet for CCRL 40/15

http://tlcv.net/16071

[Modern Times]

1 core bench is USELESS. Only for checking compile.

Well I did give 12 thread numbers too.

[Werewolf]

1 core bench is USELESS. Only for checking compile.

Not if you use Aquarium

[Jouni]

Had they changed default bench positions from SF14?

[Jouni]

Large pages didn't activate in bench. Even with LP slowdown from 14.1 is about 15%. BTW SF15 is stunning in test suites like Arasan. It solves 188/200 at 15s limit. Better than ShashChess .

[pohl4711]

Stockfish 15 was published about 6 months after the release of Stockfish 14.1
https://abrok.eu/stockfish/

Regression tests shows 36.69 elo progress at 60 sec +0.6 sec tc / 1 core/ 8 moves v3.pgn and 26.09 elo progress at 30 sec +0.3 sec tc / 8 cores/8 moves v3.pgn
https://github.com/glinscott/fishtest/w ... sion-Tests.

Ratinglist- and regression-testruns of Stockfish 15 finished:
https://www.sp-cc.de/

[Deberger]

Thanks Anst, I'll search for some of the other reported positions in the next few days and see how they go.

Andrew

Code: Select all

setoption name Hash value 256
position fen r1b2r1k/1p3pRp/2np4/p2N1p2/5P2/qN6/P1PQB1PP/5RK1 b - -
go depth 36

[bnst]

Thanks Anst, I'll search for some of the other reported positions in the next few days and see how they go.

Andrew

Sorry but I now did a deeper search and discovered that the problem persists in Stockfish 15.
After 3.5 minutes on a single thread on my slow computer I got :
35/64 03:32 194,331k 916k +5.85 Nf6+ Kh8 Nh4 Nxe5 Rce4 c4 Nxe8 Bxe8 Rxe5 cxb3 axb3 Kh7 Qd3+ Kg8 Nf5 Bb5 Qd2 Qd8 Qxd8+ Rxd8 Rd5 Rb8 Red1 Kf8 Nd6 Bc6 R5d2 a5 Nc4 Bb5 Rd5 Bxc4 bxc4 a4 g3 g6 Kg2 Nc5 Rd8+ Rxd8 Rxd8+ Ke7 Rd5 Nd7 Rb5 Kd6 Rb4 Kc5 Rxa4

After this SF seems to hang. I have now run it for 40 minutes with no output.
Regards
anst

[Andrew]

Thanks Anst, I'll search for some of the other reported positions in the next few days and see how they go.

Andrew

Sorry but I now did a deeper search and discovered that the problem persists in Stockfish 15.
After 3.5 minutes on a single thread on my slow computer I got :
35/64 03:32 194,331k 916k +5.85 Nf6+ Kh8 Nh4 Nxe5 Rce4 c4 Nxe8 Bxe8 Rxe5 cxb3 axb3 Kh7 Qd3+ Kg8 Nf5 Bb5 Qd2 Qd8 Qxd8+ Rxd8 Rd5 Rb8 Red1 Kf8 Nd6 Bc6 R5d2 a5 Nc4 Bb5 Rd5 Bxc4 bxc4 a4 g3 g6 Kg2 Nc5 Rd8+ Rxd8 Rxd8+ Ke7 Rd5 Nd7 Rb5 Kd6 Rb4 Kc5 Rxa4

After this SF seems to hang. I have now run it for 40 minutes with no output.
Regards
anst

The position just above posted by Deberger I can confirm freezes as well, so there's still a problem.

Andrew