Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: One more Albillo position

Author: Richard Pijl

Date: 08:06:53 07/08/03

Go up one level in this thread


On July 08, 2003 at 02:45:12, Jouni Uski wrote:

>[D]8/3R1P2/1ppP1p2/3r4/8/K7/p4k2/8 w - - bm Kb2
>
>Also this was considered 1998 as unsolvable by current hard/software. Not
>anymore:
>
>Analysis by Ruffian 1.0.5:
>
>1.Kxa2
>  +-  (7.73)   Depth: 20/45   00:05:02  305105kN
>1.Kxa2 f5 2.f8B f4 3.Be7 Ke3 4.Bd8 f3 5.Bh4 Kd3 6.Rf7 Rxd6
>  ²  (0.69)   Depth: 20/45   00:09:22  574275kN
>1.Kb2 a1Q+ 2.Kxa1 Ra5+ 3.Kb2 Rb5+ 4.Kc2 Rc5+ 5.Kd3 f5 6.Ra7 Rd5+ 7.Kc3 Rc5+
>8.Kb2
>  +-  (8.17)   Depth: 20/49   00:31:14  1977644kN
>

Another program able to do this in about half an hour is the Baron SMP (on AMD
2000+ MP (dual)):

 ply       time     nodes  score  pv

  7(13)&      0     12985    -19  a3a2 d5a5 a2b2 a5a8 d7a7 a8h8 a7a6 h8b8
  7(13).      0     14153    -19  a3a2 d5a5 a2b2 a5a8 d7a7 a8h8 a7a6 h8b8
  8(17)+      0     36592     17  a3a2 d5a5 a2b2 a5a8 d7b7 c6c5 b7b6 a8f8 b6b5
  8(17)&      0     55534    105  a3a2 d5a5 a2b3 a5a8 d7c7 b6b5 d6d7 a8h8 c7c6
f6f5
  8(17).      0     56524    105  a3a2 d5a5 a2b3 a5a8 d7c7 b6b5 d6d7 a8h8 c7c6
f6f5
  9(18)&      0    101113     93  a3a2 d5a5 a2b3 a5a8 d7c7 c6c5 c7b7 a8f8 b3c4
f6f5
  9(18).      0    101134     93  a3a2 d5a5 a2b3 a5a8 d7c7 c6c5 c7b7 a8f8 b3c4
f6f5
 10(20)+      0    164408    128  a3a2 d5d1 a2b3 d1d3 b3c2 b6b5 f7f8q d3d6
 10(20)+      0    236648    218  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 c6c5 e7e8
c5c4 b3c4 f8f7 d7d8q
 10(20)&      0    366882    459  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 b6b5 e7e8
f8f7 d7d8q c6c5
 10(20).      1    374565    459  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 b6b5 e7e8
f8f7 d7d8q c6c5
 11(21)&      1    541240    481  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 f8d8 f7f8q
d8f8 e7e8 f8e8 d7e8q
 11(21).      1    565597    481  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 f8d8 f7f8q
d8f8 e7e8 f8e8 d7e8q
 12(21)+      2    898299    516  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 f8d8 e7e8
d8d7 f7f8q b6b5 f8f6 f2g3
 12(23)&      3   1248585    596  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 b6b5 e7e8
f8f7 d7d8q f6f5 d8b6 f2g3 b6c6
 12(23).      3   1248606    596  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 b6b5 e7e8
f8f7 d7d8q f6f5 d8b6 f2g3 b6c6
 13(30)&      8   2366154    614  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 f6f5 e7e8
f8f7 d7d8q f7b7 d8d2 f2f3 d2c3 f3g4 e8g8 g4f4
 13(30).      8   2385893    614  a3a2 d5a5 a2b3 a5a8 d7e7 a8f8 d6d7 f6f5 e7e8
f8f7 d7d8q f7b7 d8d2 f2f3 d2c3 f3g4 e8g8 g4f4
 14(33)&     21   4665337    643  a3a2 d5a5 a2b3 a5a8 d7e7 a8d8 e7e8 d8d6 f7f8q
d6d3 b3b2 d3f3 f8b4 f2g2 e8g8
 14(36).     25   5034686    643  a3a2 d5a5 a2b3 a5a8 d7e7 a8d8 e7e8 d8d6 f7f8q
d6d3 b3b2 d3f3 f8b4 f2g2 e8g8
 15(35)+     52   8603832    678  a3a2 d5h5 f7f8q h5g5 a2b3 b6b5 f8f6 g5f5 f6f5
f2g2 b3a2 c6c5 f5f6 g2h2 d7d8 h2g3 a2a3 c5c4 a3a2 g3g2 a2a1 g2g3 a1a2 <rep>
 15(38)&    136  22647421    746  a3a2 d5a5 a2b3 a5a8 d7e7 f6f5 e7e8 a8a5 f7f8q
a5b5 b3c2 b5c5 c2d3 c5d5 d3c4 d5c5 c4d3 <rep>
 15(39).    152  24304641    746  a3a2 d5a5 a2b3 a5a8 d7e7 f6f5 e7e8 a8a5 f7f8q
a5b5 b3c2 b5c5 c2d3 c5d5 d3c4 d5c5 c4d3 <rep>
 16(42)&    573 111638014    751  a3a2 f6f5 d7a7 d5e5 f7f8q e5e2 a2a3 e2e3 a3b4
e3e4 b4c3 e4e3 c3b4 <rep>
 16(42).    573 111638036    751  a3a2 f6f5 d7a7 d5e5 f7f8q e5e2 a2a3 e2e3 a3b4
e3e4 b4c3 e4e3 c3b4 <rep>
 17(44)&    957 198112405    724  a3a2 f6f5 f7f8q d5a5 a2b3 a5b5 b3c3 b5c5 c3d3
c5d5 d3c2 d5c5 c2d2 c5d5 d2c1 d5c5 c1d2 <rep>
 17(47)&   2151 517077249    725  a3b2 a2a1q b2a1 d5a5 a1b1 f6f5 d7a7 a5b5 b1c2
b5c5 c2d3 b6b5 f7f8q c5d5 d3c2 c6c5 c2b2 d5d1
 17(47).   2234 530745415    725  a3b2 a2a1q b2a1 d5a5 a1b1 f6f5 d7a7 a5b5 b1c2
b5c5 c2d3 b6b5 f7f8q c5d5 d3c2 c6c5 c2b2 d5d1
 18(50)&   4615 1061703041    725  a3b2 a2a1q b2a1 d5a5 a1b1 a5b5 b1c2 b5c5 c2d3
f6f5 f7f8r c5b5 d7a7
 18(51).   5749 1307269042    725  a3b2 a2a1q b2a1 d5a5 a1b1 a5b5 b1c2 b5c5 c2d3
f6f5 f7f8r c5b5 d7a7

Interesting is that the Baron SMP had a superlinear speedup compared to the
single CPU version. A full 14 ply search took the SMP Version 25 seconds where
the single CPU version needed 104 seconds.

bye,
Richard.



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.