Author: Otello Gnaramori
Date: 02:06:24 11/18/01
Go up one level in this thread
On November 17, 2001 at 20:52:36, Jesper Antonsson wrote: >On November 17, 2001 at 18:24:20, Otello Gnaramori wrote: > >>On November 17, 2001 at 16:02:49, Joshua Lee wrote: >> >>>Even without the massive hardware and despite how much slower it would be is >>>there enough knowledge about how Deep thought/ Blue Played to re-create it? >>>How much time and money would it take? >> >>Hi Joshua, >>I think that as already stated somewhere in this forum , the D.B. algorithms , >>evals and so on are pretty standard ones and to be more clear are a bit outdated >>since that project was stopped on '97 by the IBM. >>The great "plus" of D.B. was and still is the hardware base of calculus with >>specialized chips capable to reach an outstanding value of 200 >>Megapositions/second . >> >>Kind regards, >>Otello. > >Otello, there is not enough information available for you to make such claims. Please see my answere here : http://www.icdchess.com/forums/1/message.shtml?197927 >To answer the original posters question, no, there is not enough knowledge about >DB to create a DB-emulator. If IBM made all DB information public (which would >probably involve documentation work for them), I think it would take perhaps six >months, full time for one person, to create an emulator. Agreed. >Of course, to emulate a massively parallel beast like that perfectly on a serial >machine would be almost impossible, I don't expect that even DB got repeatable >results all the time, but I think you could get close enough. The speed of the >emulator, NPS-wise, would be way below what we see in current micros because, >among other things, the eval is probably to complex to run quickly in software. > >Regards, >Jesper Agreed.
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.