Author: Chris Carson
Date: 09:38:08 01/14/00
Go up one level in this thread
On January 14, 2000 at 12:18:17, Robert Hyatt wrote: >On January 14, 2000 at 10:46:51, Chris Carson wrote: > >>And accepting your statement, then >>DB never played any program/computer in >>a public match (only in a lab, no published >>match details before hand, no preparations by >>the other programmers that I am aware of). >> >>IMHO: not a fair event. :( >> >>DB prototype came in third in a public tournament >>(in the light of day, not in a hidden room >>with no independant verification or validation). >> >>IMHO: a fair event. :) >> >>If you have other information such as match conditions, program setup, >>game results, pgn files from a public source and can show >>independant validation, please let me know. I like DB and >>DB junior and respect HSU, but I have to ask why no public tournaments like >>Rebel, Fritz, Hiarcs, Junior, ... have done (before the Kasparov >>match)? :) >> >>Best Regards, >>Chris Carson > > >Wait until Hsu's book comes out. Check the 'timing'. If they had played a >tournament with the real DB-1 or DB-2 hardware, it would have been within a >couple of weeks of the event. For the last event they were seriously giving >thought to either using the 1996 hardware or postponing, because the FAB shop >screwed up badly several times (not design problems, but mis-communication >inside the fab shop. IE Hsu wanted a 40-pin package. He got something bigger. >He wanted interlaced data lines across the chip to save space. They didn't >notice this. And even after all of that they had some ugly problems to solve >in a period of a week or so. IE when you have 32 'wires' running in parallel, >at high clock frequencies, with very close spacing, 'crosstalk' is a problem. > >They had to solve a bunch of problems with software kludges. And they were >barely ready to play Kasparov at match time, much less ready to do demos months >in advance. I was surprised how closely they cut it. I thought only us >academic types did that kind of stuff (IE Cray Blitz became parallel in the >two weeks immediately prior to the 1983 WCCC event we won... When I left for >New York (driving, about 24 hours non-stop) the thing wasn't working due to a >compiler problem. Harry was in Minneapolis working with Cray as we drove. That >was close. DB had the same sort of problems, which was surprising in a way. >But probably typical overall. :) Fair enough. I can not wait for the book, it should be great!! :) Best Regards, Chris Carson
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.