Stockfish 15 is ready

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

Moderators: hgm, Rebel, chrisw

Uri Blass
Posts: 10474
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Stockfish 15 is ready

Post by Uri Blass »

It seems that I get stuck with the following position without endgame tables even with the latest developement version of stockfish.

4r3/4b1R1/p7/1k2P3/1p6/1K6/1B6/8 w - -

Stockfish 220422 by the Stockfish developers (see AUTHORS file)
setoption name Hash value 256
position fen 4r3/4b1R1/p7/1k2P3/1p6/1K6/1B6/8 w - -
go depth 30
info string NNUE evaluation using nn-d0b74ce1e5eb.nnue enabled
...
info depth 26 seldepth 46 multipv 1 score cp -742 nodes 108251276 nps 1443831 hashfull 960 tbhits 0 time 74975 pv b3c2 a6a5 b2d4 a5a4 e5e6 b5c4 d4b2 a4a3 g7g4 c4b5 b2e5 b5a4 g4e4 e8c8 c2d3 c8c1 e5h8 c1d1 d3e2 d1h1 h8g7 a3a2 e4g4 h1h6 g4e4 h6g6 g7h8 g6g8 h8e5 a4b3 e4e3 b3c4 e3e4 c4d5
info depth 27 currmove b3c2 currmovenumber 1
Sopel
Posts: 390
Joined: Tue Oct 08, 2019 11:39 pm
Full name: Tomasz Sobczyk

Re: Stockfish 15 is ready

Post by Sopel »

Stockfish is still thinking, you're just not getting updates. There is no issue.
dangi12012 wrote:No one wants to touch anything you have posted. That proves you now have negative reputations since everyone knows already you are a forum troll.

Maybe you copied your stockfish commits from someone else too?
I will look into that.
User avatar
Deberger
Posts: 91
Joined: Sat Nov 02, 2019 6:42 pm
Full name: ɹǝƃɹǝqǝᗡ ǝɔnɹꓭ

Re: Stockfish 15 is ready

Post by Deberger »

Sopel wrote: Fri Apr 22, 2022 5:06 pm Stockfish is still thinking, you're just not getting updates. There is no issue.
Read about "Search explosion"

https://github.com/official-stockfish/S ... 1087012259
Jouni
Posts: 3428
Joined: Wed Mar 08, 2006 8:15 pm

Re: Stockfish 15 is ready

Post by Jouni »

Engine should output something every 5 minutes so we get off from these "is it crashed?" questions! Too difficult to add :lol: .
Jouni
User avatar
Deberger
Posts: 91
Joined: Sat Nov 02, 2019 6:42 pm
Full name: ɹǝƃɹǝqǝᗡ ǝɔnɹꓭ

Re: Stockfish 15 is ready

Post by Deberger »

Jouni wrote: Sat Apr 23, 2022 11:21 am Engine should output something every 5 minutes so we get off from these "is it crashed?" questions! Too difficult to add :lol: .
This would be easy to implement as an UCI option, a heartbeat to see the current time, nodes, depth, etc.

However I don't believe it's necessary. The operating system (and the loud cooling fan) tells me the CPU is running.

The problem is complicated search extensions which gain around 5 Elo, but sometimes cause search to get stuck in a very long loop. (See "Search explosion")

Here's another example which has a moderate positional evaluation. (i.e. Not an obvious win or obvious draw)

Code: Select all

(git checkout 6e0680efa0d0653686b8eb8753dc49718b95cb2b # Update default net to nn-d0b74ce1e5eb.nnue)

setoption name Hash value 256
position fen r1b2rk1/pp2bppp/1qn1p3/8/P1BP1P2/2N1QN2/1P4PP/3R1RK1 w - -
go
Uri Blass
Posts: 10474
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Stockfish 15 is ready

Post by Uri Blass »

Deberger wrote: Sat Apr 23, 2022 2:54 pm
Jouni wrote: Sat Apr 23, 2022 11:21 am Engine should output something every 5 minutes so we get off from these "is it crashed?" questions! Too difficult to add :lol: .
This would be easy to implement as an UCI option, a heartbeat to see the current time, nodes, depth, etc.

However I don't believe it's necessary. The operating system (and the loud cooling fan) tells me the CPU is running.

The problem is complicated search extensions which gain around 5 Elo, but sometimes cause search to get stuck in a very long loop. (See "Search explosion")

Here's another example which has a moderate positional evaluation. (i.e. Not an obvious win or obvious draw)

Code: Select all

(git checkout 6e0680efa0d0653686b8eb8753dc49718b95cb2b # Update default net to nn-d0b74ce1e5eb.nnue)

setoption name Hash value 256
position fen r1b2rk1/pp2bppp/1qn1p3/8/P1BP1P2/2N1QN2/1P4PP/3R1RK1 w - -
go
I think that it is better to output some chess line every 5 minutes because maybe the engine is running but get into some loop that is not about chess.
Dicaste
Posts: 144
Joined: Mon Apr 16, 2012 7:23 pm
Location: Istanbul, TURKEY

Re: Stockfish 15 is ready

Post by Dicaste »

Played 75 games against FT2 and it only lost 1 game at 1+10

[pgn][Event "DESKTOP-G5N9OPT, Blitz 1.0min+10.0sec"]
[Site "DESKTOP-G5N9OPT"]
[Date "2022.04.23"]
[Round "51"]
[White "Fat Titz 2 130122 64 BMI2 NUM"]
[Black "Stockfish 15"]
[Result "1-0"]
[ECO "B30"]
[Annotator "0.46;0.29"]
[PlyCount "153"]
[TimeControl "60+10"]

{Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz 3192 MHz W=49.7 plies; 6.153kN/s B=48.
1 plies; 6.634kN/s} 1. e4 {0.46/30 19 Both last book move} c5 {0.29/35 39} 2.
Nc3 {0.61/32 13 (Nf3)} Nc6 {0.34/32 4 (d6)} 3. Nf3 {0.69/27 4} e5 {0.47/32 10
(d6)} 4. Bc4 {0.55/29 10} d6 {0.30/32 10 (Be7)} 5. d3 {0.59/33 19 (a4)} Be7 {
0.31/28 5} 6. Nd2 {0.58/31 21 (0-0)} Nf6 {0.27/30 7 (Bg5)} 7. Nf1 {0.65/32 5}
Bg4 {0.34/34 9} 8. Qd2 {0.53/32 31 (f3)} Nd7 {0.25/31 8} 9. Nd5 {0.57/29 5} Be6
{0.20/33 24} 10. c3 {0.47/27 3} Nb6 {0.21/30 3} 11. Qe2 {0.50/27 9 (Nxb6)} O-O
{0.18/29 13} 12. h4 {0.76/29 3} Nxc4 {0.12/30 8} 13. dxc4 {0.75/29 5} Qd7 {
0.18/32 31 (f5)} 14. Nfe3 {0.61/33 31} Bd8 {0.26/28 2} 15. Qf3 {0.52/27 3 (a4)}
Ne7 {0.27/34 16 (Kh8)} 16. g4 {0.91/30 6 (a4)} b5 {0.49/27 3} 17. h5 {0.87/33
10} f6 {0.44/27 4} 18. cxb5 {0.54/31 15} Bxd5 {0.52/31 19 (c4)} 19. exd5 {
1.02/25 4} Qxb5 {0.28/28 4} 20. Nf5 {0.79/32 18 (Qe4)} Qa6 {0.51/33 24 (Rb8)}
21. b3 {0.87/29 8} c4 {0.62/28 4 (e4)} 22. b4 {1.31/26 3} Nxf5 {0.64/29 4} 23.
gxf5 {1.69/27 6} Qa4 {0.86/31 9} 24. h6 {1.57/29 4 (Rh4)} g5 {1.56/36 32} 25.
fxg6 {1.49/30 6} hxg6 {1.53/27 2} 26. h7+ {1.56/30 7} Kh8 {1.64/30 4} 27. Bh6 {
1.55/32 10} Rf7 {1.80/31 10} 28. Qg4 {1.81/33 7} g5 {1.77/32 5} 29. Qh5 {
1.98/33 8} Qd7 {1.54/34 5} 30. a4 {1.92/33 6} a6 {1.83/34 15} 31. a5 {1.87/34 7
} Rxh7 {2.00/31 8} 32. f4 {2.14/33 9} Qf7 {1.54/33 6} 33. Qxf7 {1.89/36 23}
Rxf7 {1.54/29 6} 34. fxg5 {2.11/32 10} fxg5 {1.69/30 13} 35. Bf8+ {2.19/29 6}
Kg8 {1.92/30 8} 36. Bxd6 {2.35/29 7} Rf4 {2.12/30 38 (e4)} 37. Ra2 {3.31/32 22
(Bc5)} Bf6 {2.33/23 4} 38. Re2 {3.72/26 4} Re8 {3.74/27 21} 39. Bc7 {3.87/28 6}
Rf3 {3.84/25 8 (Rf5)} 40. Rf1 {4.08/29 7} g4 {4.50/28 13} 41. Re4 {4.38/34 25}
Kf7 {4.43/25 2} 42. Rxf3 {4.62/27 4} gxf3 {4.28/26 4} 43. Kf2 {4.71/28 7} Rc8 {
5.33/29 23} 44. Bb6 {4.94/26 8} Rh8 {5.42/26 5 (Kg6)} 45. Re3 {5.11/27 7} Rh2+
{5.04/27 7 (Rh1)} 46. Kxf3 {5.39/26 7} Ke7 {5.58/28 7 (Rh6)} 47. Bc5+ {5.76/30
10 (Bc7)} Kd7 {6.12/27 11 (Kf7)} 48. Ke4 {6.39/25 8} Rh4+ {7.06/28 20} 49. Kf5
{6.46/26 7} Rf4+ {7.30/24 8} 50. Kg6 {6.66/25 1} Bd8 {7.75/26 14 (e4)} 51. Rxe5
{7.30/28 16} Rf6+ {7.56/21 2} 52. Kg7 {7.88/27 10 (Kh5)} Rf3 {8.74/26 18} 53.
Re6 {8.23/25 7} Rf5 {8.49/22 9 (Rg3+)} 54. Rd6+ {8.65/27 8} Kc8 {9.90/22 12}
55. Rxa6 {9.31/28 14 (Kg6)} Rxd5 {8.53/17 2} 56. Rd6 {9.41/29 8} Rd3 {10.66/25
18 (Rg5+)} 57. Bd4 {10.29/24 8} Bc7 {11.90/25 11 (Be7)} 58. Rf6 {12.99/27 20
(Rc6)} Bd8 {12.86/23 9 (Rg3+)} 59. Rg6 {16.23/22 6} Rd2 {13.57/23 11 (Bh4)} 60.
a6 {60.55/31 7} Re2 {16.70/20 10 (Bh4)} 61. Rc6+ {#20/44 9 (b5)} Kd7 {15.71/20
5} 62. a7 {#17/55 8 (b5)} Re8 {21.05/26 14 (Rg2+)} 63. Kf7 {#15/56 8 (b5)} Re7+
{#17/35 4} 64. Kg6 {#14/65 8} Kxc6 {#13/50 3} 65. a8=Q+ {#13/67 10} Kc7 {
#12/57 3} 66. Qd5 {#12/68 10} Kc8 {#11/65 5} 67. Bf6 {#11/70 12} Rd7 {#10/77 5}
68. Qa8+ {#10/83 11} Kc7 {#9/72 1} 69. Be5+ {#9/89 8} Kb6 {#8/88 7} 70. Qa5+ {
#8/97 8} Kc6 {#7/119 6} 71. Qc5+ {#7/127 8} Kb7 {#6/113 1} 72. Qb5+ {#6/228 8}
Bb6 {#5/245 1 (Ka7)} 73. Qxd7+ {#5/245 1} Bc7 {#4/245 0} 74. Bxc7 {#4/245 0
(Qxc7+)} Ka6 {#3/245 0} 75. Qc6+ {#3/245 0} Ka7 {#2/245 0} 76. Qb6+ {#2/245 0}
Ka8 {#1/245 0} 77. Qa6# {#1/245 0} 1-0

[/pgn]
User avatar
Deberger
Posts: 91
Joined: Sat Nov 02, 2019 6:42 pm
Full name: ɹǝƃɹǝqǝᗡ ǝɔnɹꓭ

Re: Stockfish 15 is ready

Post by Deberger »

Uri Blass wrote: Sat Apr 23, 2022 4:20 pm
I think that it is better to output some chess line every 5 minutes because maybe the engine is running but get into some loop that is not about chess.
It's possible to send "isready" every five minutes, but "readyok" doesn't reveal much.

Here's some code for a ping command which indicates some chess-related activity.

Code: Select all

@@ -269,6 +269,11 @@ void UCI::loop(int argc, char* argv[]) {
       else if (token == "position")   position(pos, is, states);
       else if (token == "ucinewgame") Search::clear();
       else if (token == "isready")    sync_cout << "readyok" << sync_endl;
+      else if (token == "ping")       sync_cout << "info"
+                                                << " nodes " << Threads.nodes_searched()
+                                                << " nps " << Threads.nodes_searched() * 1000 / Time.elapsed()
+                                                << " time " << Time.elapsed()
+                                                << sync_endl;
 
       // Additional custom non-UCI commands, mainly for debugging.
       // Do not use these commands during a search!
User avatar
Graham Banks
Posts: 42639
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Stockfish 15 is ready

Post by Graham Banks »

Stockfish 15 64-bit 4CPU Gauntlet for CCRL 40/15 is live here:

http://tlcv.net/16062
gbanksnz at gmail.com
Jouni
Posts: 3428
Joined: Wed Mar 08, 2006 8:15 pm

Re: Stockfish 15 is ready

Post by Jouni »

[fen]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 1[/fen]


1.Qxe5 fxe5 2.Rf1 Qe7 3.Bd1 b5 4.Kg2 Rc8 5.Bb3 Rc4 6.a4 a6 7.h4 Qe8 8.Rf2 Qe7 9.Rf3 Qe8 10.Ba2 Qe7 11.h5 Qe8 12.b4 Qe7 13.Bb3 Qe8 14.Kg3 gxh5 15.Bxc4 bxc4 16.Kh2
+- (1.72) Depth: 29/37 00:00:01 7284kN, tb=5

SF15 is STUNNING!
Jouni