M1 Apple Silicon for Chess?

[AlexChess]

Open source engines still not available on Mac Silicon M1:
Wasp 4.5.0 http://www.amateurschach.de/main/_wasp-beta-test.htm

Unfortunately, Wasp is not open source.

What is open source is the CT800, and I expect that to easily compile natively on M1 if you have Clang installed, using the provided Mac build script.

Thank you for the info

[AlexChess]

Open source engines still not available on Mac Silicon M1:
Wasp 4.5.0 http://www.amateurschach.de/main/_wasp-beta-test.htm

Unfortunately, Wasp is not open source.

What is open source is the CT800, and I expect that to easily compile natively on M1 if you have Clang installed, using the provided Mac build script.

Thank you for the info

Hi Ras!
I have added CT800 to my own topic about original engines, so we can talk about it there:
forum3/viewtopic.php?f=2&t=78182&p=906053#p906030

Best regards, Alex

[AlexChess]

Combusken 1.9.9_dev compiled by Marcin for mac Silicon M1 is the faster engine!!!!



Download macOS Silicon M1 binary: https://1drv.ms/u/s!AkW3Hj0Gl_ew0TYumDy ... g?e=xzFcEx

My topic about Promising 3000+ ELO engines with Combusken 1.9.9_dev:
forum3/viewtopic.php?p=906289#p906367

Ciao, Alex

[AlexChess]

Hi Acepoint,

Berserk 4.5.1 for mac M1 is strong and fast but crashes very often on BSG and I must restart the games.

4.6.0 has been released and maybe fixes also the crashes (not stronger, only FRC added)

Ciao, Alex ( I've cloned you, but if you want I will use Hallo )

[AlexChess]

Mac M1 Blitz 3 mins /all with 21 top engines 3000+ ELO
(most of them kindly compiled by Acepoint)



All 1050 PGN games:
https://onedrive.live.com/?cid=B0F79706 ... 69&o=OneUp
Stockfish 14 and GPLv3 compliant derivatives easily lead the tournament!
Combusken 1.9.9-dev and Zahak 6.2 new chess stars!
Sadly, Ethereal 13.00 github regression over 12.75 confirmed also on Silicon M1 Native 4 CPU, no improvements after 1 year without NNUE

[RubiChess]

Bluefish (what ever that is) draws Sf14dev and Cfish but loses 9 of 10 games against the two weakest engines. Really serious testing that you are doing. Maybe you should stop talking about regression of engines.

[AlexChess]

Bluefish (what ever that is) draws Sf14dev and Cfish but loses 9 of 10 games against the two weakest engines. Really serious testing that you are doing. Maybe you should stop talking about regression of engines.

Andreas, I only play games. Regression of Ethereal is also confirmed by all rating lists... 12.75 is simply stronger than 13.00 github (no NNUE). Look at Zahak games against Bluefish: no errors, crashes or timeouts...

Bluefish is v13.1 by MikeB... I have also watched the games against Zahak 6.0 because I was surprised too. https://onedrive.live.com/?cid=B0F79706 ... 69&o=OneUp

Thanks for you comment,
Regards, Alex

[RubiChess]

I had a look at your pgn and found only one major problem that leads to strange results like I described:
"No opening book" in your tournament describtion doesn't only mean that the engines are not allowed to use one. It also means that you play every single game from the starting position.

Some may say that this is real chess and the way to test engines. But we can see that this obviously leads to scores like

5% against 3150Elo engines
50% score against 3600Elo engines

for the same single engine.
Of course this is also sss but I wouldn't be surprised if this holds for many more games all played from starting position.

Regards, Andreas

[Guenther]

I had a look at your pgn and found only one major problem that leads to strange results like I described:
"No opening book" in your tournament describtion doesn't only mean that the engines are not allowed to use one. It also means that you play every single game from the starting position.

Some may say that this is real chess and the way to test engines. But we can see that this obviously leads to scores like

5% against 3150Elo engines
50% score against 3600Elo engines

for the same single engine.
Of course this is also sss but I wouldn't be surprised if this holds for many more games all played from starting position.

Regards, Andreas

Well I looked at it too and found a lot of serious problems, so much that I don't want to waste any more time on this.

Depths are wildly fluctuating between games and also sometimes inside the same game (for the same programs ofc).

Just one example for inside the same game
(for comparing between games I need to do some stat/calculations, but for now just trust me ;))

Code: Select all

[Event "All engines SuperBlitz 3+3"]
[Date "2021.09.26"]
[Round "1"]
[White "Zahak 6.2"]
[Black "Fat Fritz 2"]
[Result "0-1"]
[TimeControl "60+1"]
[Time "20:59:25"]
[Board "24"]
[Termination "adjudication by engines' scores"]
[ECO "B30"]
[Opening "Sicilian"]

1. e4 {+0.36/25 19.457} c5 {-0.24/33 18.087} 
2. Nf3 {+0.26/25 39.962} Nc6 {-0.32/30 2.186} 
3. Bb5 {+0.18/21 6.882} e6 {-0.41/33 21.004} 
4. O-O {+0.37/25 10.548} Nge7 {-0.32/27 1.554} 
5. c3 {+0.45/9 10.630} 

Bluefish often played with an opening book, while all other games are w/o book ;-)

Just one example here:

Code: Select all

[Event "All engines SuperBlitz 3+3"]
[Date "2021.09.26"]
[Round "1"]
[White "Crystal 3.2 130721"]
[Black "Bluefish v13.1b"]
[Result "1/2-1/2"]
[TimeControl "60+1"]
[Time "20:59:26"]
[Board "1036"]
[Termination "adjudication by egtb"]
[ECO "C83"]
[Opening "Ruy Lopez"]

1. e4 {+0.44/18 5.564} e5
2. Nf3 {+0.48/17 0.543} Nc6
3. Bb5 {+0.46/17 1.041} a6
4. Ba4 {+0.50/16 1.013} Nf6
5. O-O {+0.61/17 1.335} Nxe4
6. d4 {+0.72/17 1.237} b5
7. Bb3 {+0.55/18 1.647} d5
8. dxe5 {+0.53/19 1.955} Be6
9. c3 {+0.53/19 6.030} Be7
10. Nbd2 {+0.78/19 4.116} Nc5
11. Bc2 {+0.62/22 5.231} d4
12. Nb3 {+0.62/22 3.973} d3
13. Bb1 {+0.62/20 1.348} Nxb3
14. axb3 {+0.55/22 1.607} Bf5
15. b4 {+0.65/20 6.769} O-O
16. Re1 {+0.54/21 7.590} a5
17. bxa5 {+0.49/20 0.939} Rxa5
18. Rxa5 {+0.47/22 1.105} Nxa5 {+0.15/24 1.250} 
19. Nd4 {+0.43/22 1.537} Bg6 {+0.15/27 1.868} 
20. Bxd3 {+0.38/22 1.204} Bxd3 {+0.16/26 2.291} 
21. Qxd3 {+0.39/22 2.905} c5 {+0.24/28 3.653} 
22. Qxb5 {+0.39/22 1.236} cxd4 {+0.18/28 1.722} 
23. Bd2 {+0.37/21 1.419} Qb8 {+0.17/28 3.589} 
24. Qxa5 {+0.36/21 1.212} Qxb2 {+0.17/30 2.665} 
25. cxd4 {+0.35/21 2.435} Qxd4 {+0.16/29 2.005} 
26. g3 {+0.34/22 7.398} h5 {+0.09/25 2.634} 
27. Bf4 {+0.33/20 1.213} g5 {+0.12/29 5.178} 
28. Be3 {+0.46/17 1.659} Qd7 {+0.11/27 1.362} 
29. Rc1 {+0.49/19 4.614} Rc8 {+0.11/31 1.770} 
30. Rxc8+ {+0.32/16 0.879} Qxc8 {+0.12/33 2.152} 
31. e6 {+0.34/19 1.165} Qxe6 {+0.08/33 6.112} 
32. Bxg5 {+0.38/22 0.840} h4 {+0.08/33 1.795} 
33. Bxe7 {+0.38/22 0.881} Qxe7 {+0.06/34 1.848} 
34. gxh4 {+0.38/25 1.752} Qxh4 {+0.06/43 2.636} 
35. Qc3 {+0.38/23 0.686} Qh6 {+0.06/43 1.797} 
36. Kg2 {+0.38/24 3.512} Kf8 {+0.06/39 2.242} 
37. Qb4+ {+0.38/22 0.587} Kg7 {+0.06/42 1.886} 
38. Qd4+ {+0.37/23 1.991} Kf8 {+0.06/41 1.839} 
39. Qb4+ {+0.33/19 0.416} Kg7 {+0.06/43 1.898} 
40. Kg3 {+0.31/24 1.987} Qc6 {+0.06/42 2.939} 
41. Qb2+ {+0.26/23 0.973} Kg8 {+0.06/46 2.368} 
42. Qd4 {+0.23/21 1.065} Qh6 {+0.06/41 2.765} 
43. Qg4+ {+0.19/21 0.978} Kf8 {+0.06/37 3.076} 
44. Qb4+ {+0.16/22 0.804} Kg8 {+0.06/40 3.990} 
45. Qg4+ {+0.17/22 1.780} Kf8 {+0.05/40 1.582} 
46. Qb4+ {+0.11/21 0.626} Kg8 {+0.05/42 5.091} 
47. Qc3 {+0.07/21 1.952} Qg6+ {+0.05/37 1.390} 
48. Kf3 {+0.05/20 0.908} Qd6 {+0.05/42 1.737} 
49. Kg2 {+0.08/21 0.735} Kf8 {+0.05/41 2.215} 
50. h3 {+0.01/22 0.595} Qd5+ {+0.05/42 2.265} 
51. Qf3 {+0.00/24 2.543} Qd6 {+0.05/43 2.075} 
52. Qe4 {+0.00/23 0.347} Qb6 {+0.05/44 1.984} 
53. Qa8+ {+0.00/26 3.264} Kg7 {+0.05/44 4.061} 
54. Qa4 {+0.00/21 0.127} Qb7+ {+0.05/37 1.584} 
55. Kg1 {+0.00/19 0.395} Qb1+ {+0.05/40 1.760} 
56. Kg2 {+0.00/21 0.348} Qb7+ {+0.05/38 1.795} 
57. Kg3 {+0.00/25 1.337} Qh1 {+0.05/40 1.201} 
58. Qd4+ {+0.00/21 0.613} Kg6 {+0.05/38 1.540} 
59. Qe3 {+0.00/21 0.533} Kg7 {+0.05/38 1.471} 
60. Qg5+ {+0.00/23 0.858} Kh7 {+0.05/37 1.398} 
61. Qf6 {+0.00/22 0.812} Kg8 {+0.05/42 1.501} 
62. Kh4 {+0.00/24 3.829} Qf1 {+0.05/40 1.205} 
63. f3 {+0.00/24 0.985} Qh1 {+0.00/33 1.270} 
64. f4 {+0.00/22 0.175} Qh2 {+0.00/34 1.223} 
65. f5 {+0.00/23 0.425} Qf4+ {+0.00/40 1.259} 
66. Kh5 {+0.00/21 0.045} Qf3+ {+0.00/44 1.032} 
67. Kh4 {+0.00/31 0.357} Qf2+ {+0.00/40 1.268} 
68. Kg4 {+0.00/30 0.418} Qg2+ {+0.00/36 1.258} 
69. Kf4 {+0.00/34 0.801} Qxh3 {+0.00/37 1.710} 
70. Qd8+ {+0.00/27 0.580}
1/2-1/2

The time stamp says 1022 games are running at the same second ;-)
(this could be a Banksia bug for the timestamp pgn tag though, but who knows - I don't trust it anyway, because it went downhill
for efficiency and stability soon after the beginning)

[RubiChess]

Well I looked at it too and found a lot of serious problems, so much that I don't want to waste any more time on this.

Depths are wildly fluctuating between games and also sometimes inside the same game (for the same programs).

Just one example for inside the same game
(for comparing between games I need to do some stat/calculations, but for now just trust me )

You are probably more experienced in investigating PGNs and I trust you definitely.
I just looked at 2-3 games Bluefish vs. Zahak and they didn't use opening book.
So it is even less surprising that Bluefish results show such a big variance if it is allowed to use opening book in some games and not allowed in others.

Regards, Andreas