Author: leonid
Date: 14:06:42 01/27/02
Go up one level in this thread
On January 27, 2002 at 09:02:31, Heiner Marxen wrote:
>On January 26, 2002 at 07:11:47, leonid wrote:
>
>>Hi!
>>
>>This position is one between few hundereds that was created long time ago (June
>>2001) and never came here. Mate is very moderate and it will take not big effort
>>to solve it.
>>
>>[D]1q1kq1q1/qQrbbrQq/3QQ3/NqQqqQqN/8/BnR2RnB/3PP3/3K4 w - -
>>
>>Please indicate your result.
>>
>>Thanks,
>>Leonid.
>
>Since black is threatening Qgxd2# the search tree is small. Chest on an
>Athlon 1500+ needs just 154 second (2.6 minutes) to find a unique solution
>for mate in 9 (128 MB hash):
>
>PV: Qdxe7+ Qexe7 Qcxe7+ Qxe7 Bxe7+ Rxe7 Qff8+ Qxf8 Rxf8+ Re8 Rxe8+ Bxe8 Qec8+
>Qxc8 Qxc8+ Rxc8 Rxc8#
>
>Timing and EBF:
># 3 0.01s 1kN [ 13.35] 0.95 94- 0
># 4 0.06s [ 6.00] 10kN [ 7.28] 1.16 420- 0
># 5 0.33s [ 5.50] 56kN [ 5.63] 1.49 2267- 0
># 6 1.59s [ 4.82] 264kN [ 4.72] 1.94 11724- 0
># 7 6.92s [ 4.35] 1115kN [ 4.23] 2.56 57089- 0
># 8 30.50s [ 4.41] 4626kN [ 4.15] 3.26 268493- 0
># 9 154.58s [ 5.07] 22281kN [ 4.82] 3.86 1582341- 4173
Hi, Heiner!
Probably no other program, beyond your, can even try this position by brute
force.
I actually don't know exactly my time for 9 moves for brute force. I see only
search done by brute force in June 25. Later my program was changed. In June
time was:
Move Time Branching factor NPS.
4 0.71 sec 212k
6.92
5 4.94 sec 142k
5.87
6 29 sec 117k
5.98
7 2 min 53 sec 91k
5.46
8 15 min 46 sec 84k
Mate in 9 moves was found by selective.
Seeing your time is always interesting. It indicate, at least to me, that on
this position mine (if used all advances that you already have in your 3 ply)
solver could be speeded as much as 30 times. Around 10 times when I look speed
for 4 ply, and other 3 because of Assembler.
Cheers,
Leonid.
>Cheers,
>Heiner
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.