Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz Demonstrates World Championship Level Play at Game/30!!!!!

Author: James T. Walker

Date: 18:45:54 07/04/99

Go up one level in this thread


On July 04, 1999 at 12:32:36, Sarah Bird wrote:

>On July 04, 1999 at 12:20:33, James T. Walker wrote:
>
>>On July 01, 1999 at 17:23:03, KarinsDad wrote:
>>
>>>On July 01, 1999 at 16:50:59, odell hall wrote:
>>>
>>>>
>>>>On July 01, 1999 at 15:21:02, James Robertson wrote:
>>>>
>>>>>On July 01, 1999 at 13:43:04, odell hall wrote:
>>>>>
>>>>>>This is a weird newgroup, !! The Biggest event in computer chess history just
>>>>>>took place, yet everyone is silent!! Wake up people!  Promgrammers should be
>>>>>>celebrating at their achievement. I am not even a programmer but I am excited.
>>>>>
>>>>>G/25. And there is no World Championship at G/25. And we cannot deduce how Fritz
>>>>>would fare against Kasparov, Anand, or Kramnik; they are all 50-150 points
>>>>>higher than the masters. And not a SINGLE world championship finalist is playing
>>>>>in the masters division!
>>>>>
>>>>>James
>>>>
>>>>
>>>> Ok james I hear what you are saying! But you have to admit this is quite a
>>>>sensational result for fritz, something that cannot be ingored, when
>>>>consideredring the strength of programs at this time control.
>>>
>>>The following is what Fritz is playing on. Since Siemens is sponsering the
>>>event, you can bet that it is the top end of this and not the bottom end. If so,
>>>8 processors will make this tougher than the Fritz that played at Paderborn
>>>(which I believe was on 4 processors). If you take that into account, it is
>>>probably equilvalent to running Fritz on a single processor system at about G150
>>>instead of G25. Effectively, the increases in speed due to the multi-processor
>>>systems give the programs more "time" to search, and so they are effectively
>>>playing at slower tournament times (relative to those same programs on single
>>>processor systems). No wonder Fritz is dropping the GMs.
>>>
>>>
>>>
>>>System board
>>>Microprocessor Intel Pentium II Xeon
>>>PRIMERGY 870-40 max. 4 processors,
>>>PRIMERGY 870-80 max. 8 processors (from end of 98)
>>>Data width 64-bit internal and external
>>>Clock rate cache-bus/host bus 400/100, 450/100 MHz
>>>Second-level cache integrated 512 Kbyte, 1 Mbyte or
>>>2 Mbyte (ab 450 MHz) with ECC
>>>RAM with EDC 128 Mbyte to max. 8 Gbyte
>>>Flash-EPROM BIOS update with FD or LAN possible
>>>
>><snip>
>>>KarinsDad :)
>>
>>Hello KarinsDad,
>>The number of processores Fritz is playing on is pure speculation.  It was more
>>than a year ago that Fritz played on a dual processor system but the number of
>>processors it's using this year seems to be a big secret.  I have asked around
>>but so far no answers.  Also, it looks like Fritz may have a bug in the
>>multiprocessor operation.  The move vs Topalov 63. a7 was a definite blunder.
>>It turned a victory into a draw.  Since 63 Re8 saves the win and my Fritz 5.32
>>plays it instantly and never lets go I must assume it's either a bug in Fritz 6
>>parallel code or some of the new "Knowledge" they gave it is faulty.  So, in my
>>opinion the parallel version of Fritz6 is still in developement stages and not
>>ready.  In spite of this it's performance is impressive.
>>Jim Walker
>
>Very nice Jim, first i have heard of this position. This is a very major
>bblunder, i bet Topolov couldn't believe his luck. After less than a minute my
>5.32 sees a7 as a draw and both Re8 and Rd8 as clear wins.
>Sarah.

Hello Sarah,
I can't take credit for this.  Bob Hyatt and I was discussing where the error
was since the score went from +3 to 0.  He suggested a7 and a little analysis
shows 63. a7 is a blunder.  It's interesting to note that after 63. Re8 Re4 64.
RxR KxR 65. a7 e1=Q 66. a8=Q almost all forced, if you remove the two white
kingside pawns the tablebase instantly calls a "mt in 28".  This is all from
memory so may be an error in there somewhere.  But it looks like it's impossible
to lose this position after 63. Re8.  I suspect Topalov was on the verge of
resigning in a couple more moves.  So I'm sure he was happy with his good
fortune.
Jim Walker



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.