Author: Chris Carson
Date: 05:23:50 05/03/99
Go up one level in this thread
On May 02, 1999 at 15:34:03, Roger D Davis wrote: >Say that Hsu is able to make a Deep Blue Junior card available for the average >PC-compatible computer for $200 (my assumption, not published info). Wouldn't >that destroy the market for the commericial programs? I mean, who is going to >want to pay $100 for Millenium or even $50 for Fritz when they can pay less than >the price of Chessbase and have something that could shred all but the top ten >grandmasters (and maybe those two) on their desktop? Since a single chess chip >slowed way, way down thumped two of the leading commercials, seems that this >will destory the careers of a lot of good programmers. Either that, or you'll be >able to buy Rebel for $20, and free shipping. > >Roger I think the top programs will survive if this were to really happen, a few years ago, the top machines were the stand alone boxes (Fidelity, Mephisto, ...). The cost to develop and update a box is high, so the cost of the box is high. Software is a lot faster to change and upgrade. Advantages for DB Jr: 1. Using current chip processes (faster chess accelorators) DB Jr would be a very fast/strong program. 2. Most chess players would want it just because it is strong. 3. Raise the level of commercial computer chess. Disadvantages: 1. Expensive. My guess is it would be $1000+, to manufacture, promote, sell will be expensive and the market is not that big. 2. Would you need the latest Windows/Intel box to plug into, very expensive to create more than one platform to support. What about Mac, Unix, ... 3. DB Jr has a crude user interface. To design a Gui and a DataBase plus other features of current programs would be expensive. 4. Commercial programs would adapt quickly to DB Jr. DB Jr would be a good bet to make the top of the SSDF, but for how long and by how much? It would begin to loose points with each new release of the Commercial programs and each new release of Intel chips, not to mention that parallel programs would gain even quicker. 5. Commercial programs can also incorporate accelerator chips. Application Specific Integrated Circuits (ASIC's) can be created for the current commercial programs. One idea might be to make Move Generation an ASIC option. One chip might not be to expensive to make and could be sold to a know user group. I am sure I have missed something, but my opinion is that DB Jr would be an assest to Computer Chess and raise the bar, but I also believe we have some very smart programmers out there that would adapt and meet the challenge. 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.