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.