Houdini 1.02

[John Blake]

Robert Houdart has released Houdini 1.02

http://www.cruxis.com/chess/houdini.htm

Thanks Robert

[Robert Houdart]

You're welcome, John.
You can find more detailed information in this thread.

Robert

[Eduard Nemeth]

Thank you Robert!

ED.

[Swaminathan]

You're welcome, John.
You can find more detailed information in this thread.

Robert

Those are indeed awesome results. I'd be curious to know how it does against Rybka 4. I hope that someone runs the test.

[beram]

You're welcome, John.
You can find more detailed information in this thread.

Robert

Those are indeed awesome results. I'd be curious to know how it does against Rybka 4. I hope that someone runs the test.

Code: Select all

I am running a 50 games test against DR4 right now.
Conditions: AMD Phenom II X6 1090T 3200 Mhz,F12 gui (Fritzmark 23,1)
TC Blitz 2m+2s, 512 Mb Hash, 50 privatbook positions, ponder off

After 20 games +5 -4 = 11 for Houdini 1.02

I will update this later
Kind Regards Bram

[Swaminathan]

You're welcome, John.
You can find more detailed information in this thread.

Robert

Those are indeed awesome results. I'd be curious to know how it does against Rybka 4. I hope that someone runs the test.

Code: Select all

I am running a 50 games test against DR4 right now.
Conditions: AMD Phenom II X6 1090T 3200 Mhz,F12 gui (Fritzmark 23,1)
TC Blitz 2m+2s, 512 Mb Hash, 50 privatbook positions, ponder off

After 20 games +5 -4 = 11 for Houdini 1.02

I will update this later
Kind Regards Bram

Thanks, Looking forward to it!

[beram]

You're welcome, John.
You can find more detailed information in this thread.

Robert

Those are indeed awesome results. I'd be curious to know how it does against Rybka 4. I hope that someone runs the test.

Code: Select all

I am running a 50 games test against DR4 right now.
Conditions: AMD Phenom II X6 1090T 3200 Mhz,F12 gui (Fritzmark 23,1)
TC Blitz 2m+2s, 512 Mb Hash, 50 privatbook positions, ponder off

After 20 games +5 -4 = 11 for Houdini 1.02

I will update this later
Kind Regards Bram

Match ended in a (for me) surprising win for Houdini 1.02

Code: Select all

AMD Phenom II X6 1090T 3200 Mhz,(Fritzmark 23,1)
TC Blitz 2m+2s, 512 Mb Hash, 50 privatbook positions
                                   
1   Houdini 1.02 x64 POPCNT 8_CPU  +12/=29/-9 53.00%   26.5/50
2   Deep Rybka 4 SSE42 x64         +9/=29/-12 47.00%   23.5/50

I am letting it run a second 50 games match with Nunn2 positions
grts Bram

[beram]

an example of a black Houdini 1.02 win

[Event "Bram-PC, Blitz 2m+2s"]
[Site "AMD X6 1090T 3200 Mhz, (Fritz"]
[Date "2010.06.19"]
[Round "31"]
[White "Deep Rybka 4 SSE42 x64"]
[Black "Houdini 1.02 x64 POPCNT 8_CPU"]
[Result "0-1"]
[ECO "E90"]
[Annotator "0.38;0.12"]
[PlyCount "132"]
[TimeControl "120+2"]

{AMD Phenom(tm) II X6 1090T Processor 3207 MHz W=16.0 plies; 453kN/s B=19.3
plies; 7.757kN/s} 1. Nf3 Nf6 2. d4 g6 3. c4 Bg7 4. Nc3 d6 5. e4 O-O 6. h3 e5 7.
d5 Na6 8. g4 Nc5 9. Qc2 a5 10. Be3 h5 11. g5 Ne8 12. Rg1 {Both last book move}
Bd7 {[%eval 12,18] [%emt 0:00:05]} 13. O-O-O {[%eval 38,19] [%emt 0:00:19]} b6
{[%eval 21,21] [%emt 0:00:13] (a4)} 14. Ne1 {[%eval 53,16] [%emt 0:00:13]} c6 {
[%eval 24,21] [%emt 0:00:05]} 15. f3 {[%eval 50,17] [%emt 0:00:09]} cxd5 {
[%eval 24,20] [%emt 0:00:04] (Qe7)} 16. Nxd5 {[%eval 53,18] [%emt 0:00:07]} Ba4
{[%eval 23,20] [%emt 0:00:04]} 17. b3 {[%eval 53,18] [%emt 0:00:02]} Bc6 {
[%eval 20,20] [%emt 0:00:14]} 18. Qd2 {[%eval 53,16] [%emt 0:00:09]} Bxd5 {
[%eval 18,20] [%emt 0:00:03] (Rb8)} 19. Qxd5 {[%eval 55,16] [%emt 0:00:05]} Rc8
{[%eval 19,20] [%emt 0:00:04]} 20. h4 {[%eval 55,16] [%emt 0:00:03] (Kb1)} Qd7
{[%eval 19,19] [%emt 0:00:04]} 21. Nd3 {[%eval 49,17] [%emt 0:00:05] (Kb1)} Qa7
{[%eval 22,20] [%emt 0:00:05]} 22. Nxc5 {[%eval 49,17] [%emt 0:00:02]} bxc5 {
[%eval 16,21] [%emt 0:00:04]} 23. Bh3 {[%eval 49,17] [%emt 0:00:05]} Rc7 {
[%eval 11,20] [%emt 0:00:07]} 24. Rg2 {[%eval 45,16] [%emt 0:00:07] (Rge1)} Qa6
{[%eval 9,20] [%emt 0:00:05]} 25. Kb1 {[%eval 39,16] [%emt 0:00:04] (Qd3)} Rc6
{[%eval 10,22] [%emt 0:00:06]} 26. Bd7 {[%eval 45,17] [%emt 0:00:04]} Rb6 {
[%eval 10,20] [%emt 0:00:00]} 27. Ba4 {[%eval 44,19] [%emt 0:00:07] (Bb5)} Nc7
{[%eval 7,22] [%emt 0:00:09]} 28. Qd3 {[%eval 43,18] [%emt 0:00:04]} Ne6 {
[%eval 5,21] [%emt 0:00:06]} 29. Rf2 {[%eval 43,17] [%emt 0:00:07] (Rb2)} Ra8 {
[%eval 9,21] [%emt 0:00:27]} 30. Bd2 {[%eval 43,18] [%emt 0:00:04]} Nd4 {
[%eval 9,21] [%emt 0:00:03]} 31. Bc3 {[%eval 43,19] [%emt 0:00:05]} Qc8 {
[%eval 5,20] [%emt 0:00:03] (Rb7)} 32. Qf1 {[%eval 43,17] [%emt 0:00:09]} Qe6 {
[%eval 1,20] [%emt 0:00:21] (Qc7)} 33. Ka1 {[%eval 38,16] [%emt 0:00:15] (Rfd2)
} Kh7 {[%eval 1,19] [%emt 0:00:03] (Rbb8)} 34. Qg2 {[%eval 29,15] [%emt 0:00:
10]} Rf8 {[%eval 1,18] [%emt 0:00:03] (Rbb8)} 35. Rfd2 {[%eval 8,14] [%emt 0:
00:19]} f5 {[%eval -19,18] [%emt 0:00:06] (Rb7)} 36. gxf6 {[%eval 2,14] [%emt
0:00:04] (Rb2)} Rxf6 {[%eval -30,17] [%emt 0:00:08] (Qxf6)} 37. Rd3 {[%eval
-17,12] [%emt 0:00:03]} Rb8 {[%eval -28,17] [%emt 0:00:02]} 38. Rh1 {[%eval
-32,13] [%emt 0:00:03]} Rf4 {[%eval -36,18] [%emt 0:00:03] (Rbf8)} 39. Bxa5 {
[%eval -44,12] [%emt 0:00:03] (Rh3)} Rbf8 {[%eval -69,17] [%emt 0:00:07]} 40.
Kb1 {[%eval -48,14] [%emt 0:00:07] (Be1)} Rxf3 {[%eval -93,17] [%emt 0:00:02]
(Nxf3)} 41. Rxf3 {[%eval -114,12] [%emt 0:00:02]} Rxf3 {[%eval -93,16] [%emt 0:
00:00]} 42. Be1 {[%eval -124,13] [%emt 0:00:02] (Rg1)} Bh6 {[%eval -156,18]
[%emt 0:00:06] (Qg4)} 43. Bf2 {[%eval -156,14] [%emt 0:00:04]} Qg4 {[%eval
-156,16] [%emt 0:00:00] (Be3)} 44. Qxg4 {[%eval -145,12] [%emt 0:00:01]} hxg4 {
[%eval -204,19] [%emt 0:00:04]} 45. Bxd4 {[%eval -173,13] [%emt 0:00:00]} g3 {
[%eval -236,19] [%emt 0:00:06]} 46. Bd7 {[%eval -189,14] [%emt 0:00:02] (Rg1)}
exd4 {[%eval -230,21] [%emt 0:00:04]} 47. Bh3 {[%eval -201,15] [%emt 0:00:02]}
Re3 {[%eval -230,18] [%emt 0:00:00]} 48. Rf1 {[%eval -198,14] [%emt 0:00:02]
(Rd1)} Kg7 {[%eval -270,19] [%emt 0:00:03] (Rxe4)} 49. Rg1 {[%eval -218,14]
[%emt 0:00:04] (Bg2)} Bf4 {[%eval -279,17] [%emt 0:00:02] (Kf6)} 50. Rf1 {
[%eval -277,14] [%emt 0:00:03]} Be5 {[%eval -279,15] [%emt 0:00:00]} 51. a4 {
[%eval -331,14] [%emt 0:00:02]} Re2 {[%eval -397,18] [%emt 0:00:02]} 52. Rf3 {
[%eval -394,14] [%emt 0:00:02]} g2 {[%eval -437,18] [%emt 0:00:02]} 53. Bxg2 {
[%eval -394,15] [%emt 0:00:00]} Rxg2 {[%eval -437,18] [%emt 0:00:00]} 54. Kc1 {
[%eval -452,14] [%emt 0:00:02] (Rd3)} Rh2 {[%eval -490,19] [%emt 0:00:03]} 55.
Rf1 {[%eval -499,15] [%emt 0:00:02]} d3 {[%eval -541,20] [%emt 0:00:02] (Rxh4)}
56. Rf3 {[%eval -502,16] [%emt 0:00:01]} Rc2+ {[%eval -541,19] [%emt 0:00:00]}
57. Kd1 {[%eval -540,17] [%emt 0:00:01]} Rc3 {[%eval -546,20] [%emt 0:00:01]}
58. Kd2 {[%eval -561,18] [%emt 0:00:02]} Rxb3 {[%eval -559,20] [%emt 0:00:01]}
59. Rxd3 {[%eval -561,19] [%emt 0:00:01]} Rb2+ {[%eval -570,22] [%emt 0:00:01]}
60. Ke3 {[%eval -586,19] [%emt 0:00:02]} Ra2 {[%eval -589,23] [%emt 0:00:03]}
61. a5 {[%eval -596,19] [%emt 0:00:02] (Rb3)} Rxa5 {[%eval -594,23] [%emt 0:00:
02]} 62. Rb3 {[%eval -614,20] [%emt 0:00:03]} Kf6 {[%eval -620,20] [%emt 0:00:
02] (Ra4)} 63. Ke2 {[%eval -638,19] [%emt 0:00:04] (Kd3)} Ra2+ {[%eval -654,20]
[%emt 0:00:03]} 64. Ke3 {[%eval -654,19] [%emt 0:00:03] (Kd3)} Rh2 {[%eval
-711,20] [%emt 0:00:03]} 65. Rb8 {[%eval -654,18] [%emt 0:00:02] (Rb1)} Rh3+ {
[%eval -724,20] [%emt 0:00:03]} 66. Kd2 {[%eval -708,18] [%emt 0:00:02] (Ke2)}
Rxh4 {[%eval -724,17] [%emt 0:00:00]} 0-1

[oudheusa]

You're welcome, John.
You can find more detailed information in this thread.

Robert

Thanks Robert. What does the Analysis_Contempt parameter do?

[Swaminathan]

Match ended in a (for me) surprising win for Houdini 1.02

Code: Select all

AMD Phenom II X6 1090T 3200 Mhz,(Fritzmark 23,1)
TC Blitz 2m+2s, 512 Mb Hash, 50 privatbook positions
                                   
1   Houdini 1.02 x64 POPCNT 8_CPU  +12/=29/-9 53.00%   26.5/50
2   Deep Rybka 4 SSE42 x64         +9/=29/-12 47.00%   23.5/50

I am letting it run a second 50 games match with Nunn2 positions
grts Bram

Results are close. Houdini is definitely getting a LOT better. Whatever the author claimed seems to be conservative estimate (a modest 20+).