Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How will engines get much better?

Author: Robert Hyatt

Date: 07:09:23 06/12/99

Go up one level in this thread


On June 11, 1999 at 23:23:25, Paul Richards wrote:

>On June 11, 1999 at 20:43:23, Robert Hyatt wrote:
>
>
>>To you, yes.  to chess in general?  No... because we have _already_ seen a
>>machine run at 200M nodes per second and beat Kasparov.  So it won't represent
>>a huge jump in general, although when compared to the micro computers/programs
>>of today, it will represent a jump of +400 Elo or more...
>
>I think it will have a substantial impact in the home preparation of
>top players, and this will have a trickle down effect on the popularity
>of particular opening variations.



Sure.. but we are apparently having a slight grammatical misunderstanding.

The original discussion was "will computer chess make any substantial leaps
forward?"  and I responded _no_.  And from my perspective, that was a correct
statement, because we _already_ have a deep blue at 200M nodes per second, and
have had one for over 2 years...  so the DB chip Hsu plans to market won't
represent any huge jump forward.  A huge jump beyond where current micro-
programs are? Sure.  But not an _overall_ huge jump in computer chess...

I think that is where the communication is breaking down.  We've had very fast
hardware chess engines since the 70's, three iterations of belle, the first
with a move generator, the second with hardware eval, the last the full-blown
search+everything belle that won the 1980 WCCC.  Then we had Greenblatt's
CHEOPS, BEBE, Deep Thought, HiTech, and Deep Blue...  The average consumer
didn't have access to such boxes of course.. but they were around in the
computer chess circles...



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.