Author: Uri Blass
Date: 08:13:05 09/07/01
Go up one level in this thread
On September 07, 2001 at 09:35:37, Slater Wold wrote: >On September 07, 2001 at 06:36:13, Uri Blass wrote: > >>On September 07, 2001 at 05:39:59, Slater Wold wrote: >> >>>On September 07, 2001 at 02:59:35, Gregor Overney wrote: >>> >>>>Although, Kramnik's level compares with Kasparov's level, deep fritz is much >>>>worse than IBM's deep blue from 1997. This is nothing more than an interesting >>>>game between a top GM vs. a smaller computer system (eight CPUs running at 900 >>>>MHz?). But it cannot compare with the event from 97 in New York. >>>> >>>>Deep fritz will not win this match. It might draw a few games but I would be >>>>surprised if deep fritz can win one game. >>>> >>>>According to http://www.chessbase.com/events/bahrain/sched1.htm, Kramnik gets >>>>plenty of rest between games. Something, Kasparov could not enjoy in New York. >>>> >>>>Gregor >>> >>>Just about everyone in this forum knows Deep Fritz is going to get crushed. >> >>I am not sure that everybody in this forum knows it but I am almost sure that >>Deep Fritz is going to get crushed. > >Me either. Hence the usage of "Just about everyone........." > >> >>My prediction after hearing that kramnik got the program is 100% for kramnik. > >Yes. Kasparov did not get a personal copy of Deeper Blue. :) > >> >>>Everyone in this forum knows that Deep Fritz on 8x900mhz is _NOTHING_ compared >>>to Deeper Blue. >> >>Wrong. >> >>Not everybody knows that deep Fritz is nothing compared to Deeper blue. > >If you cannot tell the computing power difference between 1000 processors, and a >"supercomputer" vs. an 8 way 900mhz box, you need help. The computing difference is huge but making it work efficiently is a very hard task and I doubt if IBM had the time to do it. It is impossible to compare only by nodes. Deeper blue did not use null move pruning when Deep Fritz does it. I agree that Deeper blue was the best at 1997 but programmers learned a lot after 1997. > >> >>I do not know it,Amir Ban does not know it,Vincent dipeveen does not know it. >>I believe that a lot of other people also do not know it. > >I know that Deeper Blue was seeing 200M nodes a second. I know that, based on >calculation, Deep Fritz will be seeing 4M - 6M nodes a second. Well, I believe >you can take the _worst_ C++ chess playing program in the world, and run it on a >4M - 6M node per second machine, and then run it on a 200M node per second >machine, and you're going to get a significant increase in performance. Hsu and >Murray were _NOT_ idiots. They did _NOT_ have an amateur chess program. This >program was more refined, and had more money and resources put into it than Deep >Fritz ever will. If you cannot see where this might make a difference, I am >sorry. I have seen Bob try to explain it to you several times, and your >obviously not going to listen, so I am not going to waste my breath here. > >> >> >>> >>>It's publicity. And that's all it is. >>> >>>The world will soon know the difference between "Chessbase" and "IBM". >> >>I am not going to know it. > >Obviously. > >>The match is not played under similiar conditions so comparison is useless if >>Fritz does not win. > >I am _NOT_ the one comparing it! The press is! I said there was _NO_ >comparison, REMEMBER?! > >> >>If Deep Fritz wins then it means that Deep Fritz is better but if Deep Fritz is >>crushed it means nothing. >> >>I have an example when Deep Fritz can find the right move(44.Kh1 in game 2) when >>Deeper blue could not see it. >> >>I still wait to see examples when it is the opposite and Deep Fritz need many >>hours to find what Deeper found in 3 minutes. > >LOL - This is funny. I can give you at least one a game. You need only to give me moves of Deeper blue that Deep Fritz needs hours to find or moves when there is no doubt that deeper blue found the right moves and Deep Fritz cannot do it. > >> >>If you can give me a lot of examples I may change my mind. > >No you won't. I can change my mind when I see new facts. Uri
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.