Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: how to beat krammnik?

Author: woehrle

Date: 13:54:07 11/07/00

Go up one level in this thread


On November 07, 2000 at 15:08:02, Uri Blass wrote:

>On November 07, 2000 at 11:19:19, woehrle wrote:
>
>>I have a question to the programmers of chess programs
>>
>>I think all of you know SETI@home, a scientific experiment, that use home-
>>computers to look for extraterrestrial intelligence (SETI = Search for
>>Extraterrestrial Intelligence). At the moment more than 4 million home-pc are
>>connected.  A gigantic performance. Question: it is possible to play a Game
>>"World against Kramnik". by combining  50.000 - 500.000 home-pc.? I think, that
>>krammik would have no chance?!
>
>1)I think kramnik has good chances in this game.
>I even doubt if being 10000000000000000000 times faster is going to help to get
>more than a draw against kramnik.
>
>In order to win kramnik you need to cause him to do errors and playing perfect
>may be not enough.
>
>2)It is theoretically possible to combine 50000-500000 home pc but there should
>be clear rules how to use them.
>I have no idea how to use the alpha beta with 50000-500000 home pc but it is
>possible to get 3 plies deeper without using the alpha beta for parallel search.
>
>All you need to do is to give the 50000-500000 computers to calculate all the
>possible positions that are 3 plies deeper than the root position and calculate
>a score for the root position that will be calculated from this data.
>
>Every computer operatoer will have a fixed number and the computer will
>calculate the position that is determined by the number.
>
>After doing all this calculation the operators will send all the numbers to one
>computer operator by email and his(her) computer is going to read all the
>50000-500000 emails and calculate a move for the root position by this
>information.
>
>This idea may be productive for a correspondence game but I doubt if it can be
>productive for tournament time control game.
>
>possible problems:
>1)I do not know the time that computers need to read 50000-500000 emails.
>2)There may be connection problem of few operators and in this case they have to
>use telephone to give the information to other operators.
>
>There may be more problems that I did not consider.
>
>Uri

 don't agree.The present fastest  computer of the world "ASCI-White" is 1000
times faster - 12,3 teraflop-   than the computer Deep Blue", which won the
match against  Kasparov 1997.
2474165 home computers of the SETI project rendered a  performance of 26.42
tera-flops in  the last 24 h see:setiathome.ssd.berkeley.  I think that this
performance is able to beat Kramnik in a competition.



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.