Rook endgame, eight cores against one Toga, almost a draw...

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

Moderator: Ras

User avatar
Eelco de Groot
Posts: 4703
Joined: Sun Mar 12, 2006 2:40 am
Full name:   Eelco de Groot

Rook endgame, eight cores against one Toga, almost a draw...

Post by Eelco de Groot »

The first day at Playchess, my poor computer had a hard time facing all those Quads and Octals and still trying to play attacking chess. Also some disastrous crashes, because of Toga using the wrong egbb.dll for accessing bitbases. But it was certainly instructive and a lot of fun, and the first real practice against Rybka 2.3.2a. I have not seen any of the new Rybka betas in action today on Playchess, but I bet they will appear, later this month for sure. What do you guys think, will Rybka 3 be ready for Leiden?

I played a few games with my brand new Hiarcs 12 SP on the server, the main reason for buying Hiarcs the Chessbase version is my interest in the new Hiarcs 12 book by Eric Hallsworth, and also believe it or not I did not yet own a Chessbase program so I thought it worth the slightly higher price for this version. From what I have seen of it thus far I can certainly recommend it! Great program Marc, Harvey, Eric and the rest of the Hiarcs team!

Played also a few other games to test Toga in this environment, using Hiarcs new book to offset the hardware difference a bit :)

My handle on Playchess is Ancalagon and as you maybe guess, that is more about dragons than about Fruit! But Ancalagon is not ready yet and I'm not quite sure yet when it will be, and whether a separate release would be acceptable. But work on this Glaurung version will certainly continue! At least until Rybka 3, trying to keep Ancalagon 2.1 within 200 points distance of Rybka 3 is the goal for the moment! Wimbledon and football and everything else will just have to wait!

Toga meanwhile had a bit of a hard time in this environment as I already described, but Hiarcs 12, even on my mediocre Athlon machine played a few very good games against super Quads with Rybka!

In one game though it was Toga Checkov that played not bad at all and resisted, until far in the endgame against a dual Xeon, 8 cores, of Ultra-D in a long 60 + 15 game. Rybka often calculated about 5 plies deeper! :wink: Probably even more in the endgame. Then Toga crashed when it tried to access bitbases in the wrong folder, but the final position was already difficult. It is possible there may have been better possibilities a few moves earlier, but Toga's endgame knowledge, purely codewise, in Rook endings is probably quite a bit below that of Rybka, not taking into account those five, six, no seven extra plies! Plus two, because of obfuscating the plydepths I strongly suspect! Bit of a steep task for Toga, but instructional it certainly was and who knows once in a hundred games like this..

[Event "Partij met rating, 60m + 15s"]
[Site "Playchess, Computers toegestaan"]
[Date "2008.05.11"]
[Round "?"]
[White "Ultra-d, Rybka 2.3.2a mp"]
[Black "Ancalagon, Toga Checkov 1.4 Beta 4"]
[Result "1-0"]
[ECO "B19"]
[WhiteElo "2578"]
[Annotator "0.09;0.51"]
[PlyCount "89"]
[EventDate "2008.05.11"]
[TimeControl "3600+15"]

{Rybka 2.3.2a mp: 22.2 ply; 749kN/s Intel(R) Xeon(R) CPU E5420 @ 2.
50GHz 3000MHz, (8 threads :!: ), Toga: Athlon 2009 MHz, MOD Hiarcs 12.ctg, 256 MB} 1. e4 {B/0 0} c6 {B/0 0} 2. d4 {
B/0 0} d5 {B/0 0} 3. Nd2 {B/0 0} dxe4 {B/0 0} 4. Nxe4 {B/0 0} Bf5 {B/0 0} 5.
Ng3 {B/0 0} Bg6 {B/0 0} 6. h4 {B/0 0} h6 {B/0 0} 7. Nf3 {B/0 0} Nd7 {B/0 0} 8.
h5 {B/0 0} Bh7 {B/0 0} 9. Bd3 {B/0 0} Bxd3 {B/0 0} 10. Qxd3 {B/0 0} Ngf6 {B/0 0
} 11. Bf4 {B/0 0} e6 {B/0 0} 12. O-O-O {B/0 0} Be7 {B/0 0} 13. Kb1 {B/0 0} Qa5
{B/0 0} 14. Ne4 {B/0 0} Nxh5 {B/0 0} 15. Bd6 {B/0 0} Nhf6 {B/0 0} 16. Bxe7 {
B/0 0} Kxe7 {B/0 0} 17. Nxf6 {B/0 0} gxf6 {B/0 0} 18. Nd2 {0.09/19 232} Rag8 {
B/0 0} 19. g3 {0.11/19 52} h5 {B/0 0} 20. Nc4 {0.15/20 111} Qb4 {B/0 0} 21. Ne3
{0.14/19 80} Kd8 {B/0 0} 22. Rh4 {0.21/18 25} Qd6 {B/0 0} 23. c4 {0.25/19 99}
Kc8 {0.51/15 162} 24. c5 {(De2) 0.13/21 217} Qc7 {0.49/16 168} 25. Nc4 {
0.00/22 145} Kb8 {0.48/16 48} 26. Rdh1 {(Pd6) 0.09/21 25} f5 {0.32/15 118} 27.
Rxh5 {0.05/24 162} Rxh5 {0.20/17 0} 28. Rxh5 {0.08/24 42} Nf6 {0.23/17 115} 29.
Rh3 {0.06/23 367} Ne4 {0.19/17 0} 30. Qe3 {0.05/23 119} b5 {0.28/16 146} 31.
Nd2 {(cxb6) 0.04/22 158} Nf6 {0.26/15 114} 32. Nf3 {(Th4) 0.06/22 44} Ng4 {
0.25/16 145} 33. Qe1 {(De2) 0.07/20 104} Ka8 {0.11/15 95} 34. Ne5 {
(Th7) 0.22/21 99} Rd8 {0.09/16 170} 35. Rh7 {0.18/25 75} Nxe5 {0.28/16 142} 36.
Qxe5 {0.17/26 46} Qxe5 {0.33/16 52} 37. dxe5 {0.11/28 12} Rd1+ {0.21/17 95} 38.
Kc2 {0.29/25 0} Rf1 {0.21/18 120} 39. f4 {0.14/26 48} Rf2+ {0.24/17 78} 40. Kc1
{(Kd1) 0.62/27 35} Rf1+ {0.24/18 87} 41. Kd2 {0.62/30 59} Rf2+ {0.25/18 55} 42.
Ke3 {(Kd1) 0.62/30 7} Rxb2 {0.00/13 26} 43. g4 {(Txf7) 0.62/29 70} fxg4 {
0.00/13 84} 44. Rxf7 {1.04/23 23} Rxa2 {0.00/16 83} 45. f5 {
Ancalagon geeft op 3.52/26 16} 1-0

11: Ultra-d,R - Ancalagon, Partij met rating, 60m + 15s 2008
[d]k7/p4R2/2p1p3/1pP1PP2/6p1/4K3/r7/8 b - - 0 1

Analysis by Hiarcs 12:

1. +- (2.63): 45...Ta3+ 46.Kd2 Ta2+ 47.Kc3 exf5 48.e6 Te2 49.e7 Te3+ 50.Kd2 f4 51.Tf8+ Kb7 52.e8D Txe8 53.Txe8 g3 54.Tg8 b4 55.Kd3 Ka6 56.Kc4 Ka5 57.Tg4
2. +- (2.65): 45...exf5 46.e6 Ta1 47.e7 Te1+ 48.Kd2

(Groningen 11.05.2008)

Eelco