Author: Chrilly Donninger
Date: 10:46:06 12/12/05
Go up one level in this thread
On December 12, 2005 at 13:16:55, Ziad Haddad wrote: >Hi Joshua >unfortunately it's only a dream, it couldn't be possible as Dr Chrilly >Dronninger said in previous posts just a few hours before. Maybe I get something wrong. Do you mean, that Rybka is implemented in an FPGA? I have not the slightest intention to do that. I wanted to see for curiosity whats going on under the Rybka-hood. Its an interesting programm. I only said, that one can not take trick X of one programm and just implement it into ones own. The parts must fit together. It is also not possible to implement any PC-Programm 1:1 in an FPGA-card. Such a card has a complete other architecture, and some things which are on the PC very efficient, are on the FPGA almost impossible and the other way round. One always has to optimize for the ressources at hand. E.g. a bitboard makes no sense for an 8-Bit chip, but it works fine for a 64-Bit architecture. Besides this: We always wanted to simulate Tal. In that way we have come far. Rybka is more Karpov than Tal. It is interesting if these two different philosophies/styles play against each other, but there is no point to reverse Tal into a Karpov. Chrilly
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.