Author: Keith Evans
Date: 15:23:36 06/01/05
Go up one level in this thread
On June 01, 2005 at 13:29:56, Daniel Staf wrote: >On May 31, 2005 at 16:57:08, Gian-Carlo Pascutto wrote: > >>On May 31, 2005 at 16:12:19, Matthew Hull wrote: >> >>>On May 31, 2005 at 15:35:30, Daniel Staf wrote: >>> >>>>Hi, >>>> >>>>Is there anyone is this forum who has experience in implementing/porting a chess >>>>ai to a microcontroller or fpga? >>>> >>>>I'm about to build a stand alone chess computer with self moving pieces (they >>>>are moved by electromagnets under the board) Currently I'm looking for some >>>>piece of code so I don't have to start from the very beginning. >>>>I have thought about implementing this in either a separate microcomputer and an >>>>fpga or maybe I can program the cpu in the fpga too. >>>> >>>>Anyone tried implementing an chess ai this way? >>> > >>> >>>I would think that using a cheap, conventional processor board and a free chess >>>engine, would get you a stronger dedicated unit than even a TASC R30, easily. >> >>Apparently it's not *that* trivial (see Resurrection module). >> >>Those old programs were damned well optimized for those small cpus. >> >>-- >>GCP > > >Thanks for the feedback :) > >I'm quite new in chess programming but I have good experience in >fpga-implementations. Some of todays fpga:s (developer environments) have built >in support for designing embedded processors. The problem is that they are quite > expensive. >If a microcomputer can make a resonable good implementation then I probably go >for that. But it's hard to know if it will run good enough whithout putting to >much time on optimization. After all they are not that fast. > >Anyone know how much memory is needed for an quite good ai? You might try looking into - http://www.gumstix.com/ I think that it will be hard to beat this in an affordable FPGA. Plus it seems like a nice development environment. If you do it well it may be an appropriate program to port to the Resurrection module. Also for FPGA stuff there's always Arm7 plus FPGA - http://www.charmedlabs.com/xportmain.htm If you really try to take advantage of the parallelism of an FPGA for chess it's not a trivial project. It might be better to start with a microprocessor to learn, and then later consider a move to a pure FPGA or hybrid approach. -Keith
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.