Author: leonid
Date: 06:31:38 06/28/00
Go up one level in this thread
On June 28, 2000 at 06:39:09, Ralf Elvsén wrote: >On June 27, 2000 at 21:24:44, leonid wrote: >> >>Tom, I am sometime in Chess Club not that much for finding how to write my >>program, not even how do it competitive. This I can do alone. This is simple. I >>am here mainly to go "gossip" a little bit about chess programming and verify >>some parameters in program. Recently it was about "effective branching factor", >>raw speed of certain part of my program, and so like. Sometime it is very easy >>to find what you are been looking for. Like explanation that I found one day >>about "extensions", when I even not knew that they existe. Sometime it is not >>so. You can talk to somebody and find more about personality of man that about >>data that you need. And what I never, ever expected is that this place is so >>serious. You must take special care for using only precise, sacred words. >>Otherwise, you will be never understood and left in silence. >> >>Leonid. > >Leonid, can you tell me how many seconds your positional logic needs to >solve the "metapositions" in a typical chessgame-position? With >"metaposition" I mean the "relevant" information in terms of >"nodeevaluations" that is "attached" to >the original position in terms of searching brute-force with no extensions. >This would be very helpful to me. > >Ralf > >(friendly) Just send to me exact position or put it here. Will say you exact NPS. If I understood your expressions correctly, you are asking me about nodes/per/second when search is done exactly like my program is doing right now. Computer AMD 400Mhz. Program use no extensions or hash tables. Brute force search. Fixed depth. When I looked into average NPS counter (it count just what you said) and my program played against itself for first 50 moves (starting with initial position), I had two times numbers: 200000 and 167000 positions/in one second. I also remember that when I looked into average NPS (around 6 months ago), for the positions that I took from some Chess Revue, it was average 210000 nodes/second (or positions/second). Recently I looked what is number of NPS for minimax. Did this around 2 weeks ago. It was between 800000 and 1100000 nodes/second. Leonid.
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.