Author: Vincent Diepeveen
Date: 16:26:12 04/09/03
Go up one level in this thread
On April 09, 2003 at 19:02:51, Keith Evans wrote: >On April 09, 2003 at 18:53:14, Vincent Diepeveen wrote: > > >>Don't they all need to get shipped by me a hardware card each time? >> >>Is there *any* way i can update them by email a new version? > >Depends on the card. Some PCI cards are configured right over the PCI bus. Many >people are using compact flash now for standalone boards. So yes - email would >work. > >>That's one speed part. second speed part is how many Mhz can we clock these >>things? > >Depends on the design - can't tell you now. Yes the parts get faster over time >for multiple reasons. More routing, better process,... > >>So it needs like 4 years to get the FPGA to double in speed. Then each 'compile' >>takes hours and hours and hours. Even a very fast machine with a lot of RAM >>isn't speeding up synthesizing. > >A 2.8 GHz Xeon can synthesize, place and route a design pretty quickly - within >hours. (Trust me - I've been there done that.) Definitely worth the money to >upgrade your computer for this. > >>However you lose a lot because of not being capable of using hashtables and the >>move ordering is a lot more poor than in software of course. > >You can of course use hashtables. Nothing to prevent that. I don't understand >why people say this. Just because nobody did it yet? Well most importantly because it makes the end product too expensive to sell. whatever you create, it must be like under $150k to produce. If it is $150k then the end user already pays more than $500 for such a card which is already in the danger zone. If it gets $200 then it sells for $750 or so. if it produces for $250 then you sell for $1000. $1000 is just too much for such cards. producing couple of thousands of 'chips' from it: $50 getting couple of thousands of pci cards + license: $100 that's already $150 so if you put for $100 ram at it with patented stuff then it's $250 which makes the card too expensive. be realistic that you must count on selling 10000 of these things. that is already a big risk! >Anyway part of this would just be for experimentation. Exploring alternatives to >the boring Von Neuman approach man! > >Regards, >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.