Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: If You want to see how 2,5 THz(!) PC play ...

Author: Joachim Rang

Date: 15:14:32 03/14/04

Go up one level in this thread


On March 14, 2004 at 12:42:13, Jouni Uski wrote:

>...
>This is only reminder for all who have missed this. In Nickel's correspondence
>match
>engines get average 2,5 days for one move = 3600 minutes = equal to 1200*3
>minutes. This means same performance as
>2,5 THz PC in 3 minutes play!! In another way 1 Gnps for Fritz/Junior!! So is
>this near
>perfect play now? In game 6 / Pluto engine is said to missed winning tactical
>16. g4!! move.
>In game 3 / Jupiter engine has played some dubious moves and is fighting for
>draw. In other
>games no fear for lose to programs I think.
>
>Jouni

yes I follow this match with great interest. Intersting observation, that this
timecontrol is equal to a 2.5 GHz - CPU and 3 minutes per move.
So far it looks like the engines do pretty well. OTOH in game 6 Pluto missed the
winning 16.g4!! which shows clearly that this isn't perfect chess. It's also
clear that Pluto isn't Junior since Junior finds this move quite quickly. In
game 3 Arno simply found a very deep tactical combination which the program was
also unable to spot. In game 1 Arno also almost outplayed the engine but with
exact play the engine was able to defend very well. But in three games (!) the
engines were able to achieve better positions - this is quite amazing
considering the fact that Arno also may use engines.

regards Joachim

http://direct.chessfriend.com/Nickel_Match/nickel_match_e.htm



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.