Author: Tord Romstad
Date: 06:59:41 08/30/03
Go up one level in this thread
On August 30, 2003 at 09:51:49, Jon Dart wrote:
>On August 30, 2003 at 07:56:28, Tord Romstad wrote:
>
>>Here, like in many tactical positions, Gothmog has a rather bad branching
>>factor. Is this normal, or should I start looking for a bug?
>
>You are actually solving this after searching about the same number of nodes as
>Arasan (10946267 is your node count at ply 9, I presume).
Yes, 10946267 is the node count when f5 appears as the first move of the PV
at ply 9. I use MTD(f), therefore I have to do several re-searches
to get the final score for the move ('<=' means 'fail low', while '>='
means 'fail high').
>The problem is that you're taking 82.6 seconds to search 10.9 million
>nodes - about 132,000 nodes/second. This is < 1/3 of Arasan's nps, so the
>solution takes longer. So either you have a slower program (takes longer
>to process each node) or a slower machine.
I forgot to add information about my hardware. The machine is slightly
slower than yours (I use a PIV 2.4 GHz), so the main explanation for
my long solution time is that my program is very much slower than yours.
132,000 nodes/second is actually a bit more than I usually get in middlegame
positions.
Tord
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.