World Computer Chess Championship 2015

[whereagles]

what exactly is this 2400 cores thing of Jonny...? did they bring the Deep Blue machine with them and plugged Jonny's algorithm in it or something??

[Sean Evans]

what exactly is this 2400 cores thing of Jonny...? did they bring the Deep Blue machine with them and plugged Jonny's algorithm in it or something??

http://www.scientific-computing.com/new ... ws_id=2172

https://chessprogramming.wikispaces.com/Jonny

[whereagles]

what exactly is this 2400 cores thing of Jonny...? did they bring the Deep Blue machine with them and plugged Jonny's algorithm in it or something??

http://www.scientific-computing.com/new ... ws_id=2172

https://chessprogramming.wikispaces.com/Jonny

Oh... they're using a major computer cluster.. lol

Well.. amazing that Komodo comes very close with a mere 1% of that gargantuan computer power.

[gerold]

what exactly is this 2400 cores thing of Jonny...? did they bring the Deep Blue machine with them and plugged Jonny's algorithm in it or something??

http://www.scientific-computing.com/new ... ws_id=2172

https://chessprogramming.wikispaces.com/Jonny

Oh... they're using a major computer cluster.. lol

Well.. amazing that Komodo comes very close with a mere 1% of that gargantuan computer power.

Its a test of compute power, Didn't you know that.

[Uri Blass]

what exactly is this 2400 cores thing of Jonny...? did they bring the Deep Blue machine with them and plugged Jonny's algorithm in it or something??

http://www.scientific-computing.com/new ... ws_id=2172

https://chessprogramming.wikispaces.com/Jonny

Oh... they're using a major computer cluster.. lol

Well.. amazing that Komodo comes very close with a mere 1% of that gargantuan computer power.

I see nothing amazing.

I think that
Komodo was the favourite to win the competition and I consider the fact that Jonny won as a surprise.

Jonny also used better hardware in the past but did not win WCCC.

Note that I still believe that Komodo 24 cores is better than Jonny 2400 cores inspite of the results.

It is not easy to make efficient use of so many cores so I think that 100:1 hardware advantage is clearly misleading.

[Jeroen]

In the WCSC Komodo has already made 3 draws in 4 rounds: against Ginkgo, Hiarcs and Shredder.

IMO this poor performance is mainly due to questionable opening choices. I wonder if they ever tested these poor lines, or just thought that any book line will do.

Same for Hiarcs, that plays +25 moves out of book and then finds itself in a totally drawn position (against Protector) and even a bad position (against Ginkgo).

Poor book choices decide these tournaments.

[Modern Times]

Poor book choices decide these tournaments.

Book choice can play a major part. Many years ago I switched to chess960 for that very reason. No more arguments about good books, bad books, long book, short books, wide books, variety etc etc

[Jeroen]

I think a 'backwardation' is in order: shorter book lines are better!

Just take the example below: Hiarcs plays about 25 moves out of book and then can only hope for a draw. Nice game by Ginkgo, no chance for black whatsoever. Such lines 'kill' your tournament....

[Event "WCSC"]
[Site "Leiden"]
[Date "2015.07.04"]
[Round "4"]
[White "Ginkgo"]
[Black "Hiarcs"]
[Result "1-0"]
[PlyCount "141"]
[TimeControl "60+1"]

1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. Bg5 h6 6. Bh4 dxc4 7. e4 g5 8. Bg3 b5 9. Be2 Bb7 10. O-O Nbd7 11. Ne5 h5 12. Nxd7 Qxd7 13. Be5 Rh6 14. Qc1 Ng4 15. Bxg4 hxg4 16. Qxg5 Rg6 17. Qh4 Be7 18. Qh8+ Bf8 19. Qh5 Qe7 20. Rfd1 O-O-O 21. Bg3 Rg5 22. Qh4 Qf6 23. Be5 Qg6 24. Qg3 Rd7 25. Qe3 Rh5 26. Ne2 Qh7 27. Bg3 f6 28. Qf4 Bd6 29. e5 fxe5 30. dxe5 Bxe5 31. Rxd7 Qxd7 32. Qxg4 Bxb2 33. Qg8+ Qd8 34. Qxe6+ Qd7 35. Qg8+ Qd8 36. Qg4+ Qd7 37. Qxh5 Bxa1 38. h4 Ba6 39. Nf4 Bd4 40. Qg6 b4 41. Ne6 Bb5 42. h5 b3 43. axb3 cxb3 44. Nxd4 Qxd4 45. Qf7 Kd8 46. Qf8+ Kd7 47. Qf5+ Ke7 48. Be5 Qd5 49. Qf6+ Kd7 50. h6 Bd3 51. Qg7+ Ke6 52. f4 Qc5+ 53. Kh2 Qe7 54. Qg3 Qh7 55. Qh4 Qd7 56. Qg4+ Bf5 57. Qg8+ Qf7 58. Qc8+ Kd5 59. Qd8+ Kc4 60. g4 Bg6 61. Qd4+ Kb5 62. f5 Bh7 63. Qd3+ Kb6 64. Kg3 c5 65. Qd6+ Kb5 66. f6 b2 67. Bxb2 Bg6 68. Qe7 Qb3+ 69. Kh4 Qc2 70. Kg5 Bh7 71. f7 1-0

[Jeroen]

Or take a look at this game:

[Event "WCSC"]
[Site "Leiden"]
[Date "2015.07.04"]
[Round "?"]
[White "Hiarcs"]
[Black "Komodo"]
[Result "1/2-1/2"]
[ECO "D27"]
[PlyCount "81"]
[EventDate "2015.07.04"]

1. d4 e6 2. c4 d5 3. Nf3 dxc4 4. e3 Nf6 5. Bxc4 c5 6. O-O a6 7. Bb3 b5 8. a4 b4 9. e4 Bb7 10. e5 Ne4 11. Nbd2 Nxd2 12. Bxd2 cxd4 13. Ng5 Be7 14. Qh5 Bxg5 15. Bxg5 Qd7 16. Rfd1 Nc6 17. Rac1 h6 18. Be3 Qe7 19. Bxd4 Nxd4 20. Rxd4 Rd8 21. Rxd8+ Qxd8 22. Bxe6 g6 23. Qh3 fxe6 24. Qxe6+ Kf8 25. Qxg6 Qd2 26. Qf5+ Ke8 27. Qe6+ Kf8 28. Qf6+ Kg8 29. Qe6+ Kf8 30. Qf6+ Kg8 31. Qg6+ Kf8 32. Qf5+ Ke8 33. Qh5+ Ke7 34. Qh4+ Ke8 35. Qh5+ Ke7 36. Qh4+ Ke8 37. Qc4 Rf8 38. Qe6+ Kd8 39.
Qb6+ Ke8 40. Qe6+ Kd8 41. Qb6+ 1/2-1/2

You cannot win tournaments by playing such lines .

[ernst]

Besides operator, isn't Erdo the book creator of Komodo a well?