CPU design at the gate level

From: Ben Franchuk <bfranchuk_at_jetnet.ab.ca>
Date: Fri Nov 2 19:22:30 2001

Richard Erlacher wrote:
>
> While it's true that TTL and work-alike CMOS is found everywhere, if you go with
> a parts list, you'll find you're unable to get a substantial range of parts
> these days. You may be able to find the parts eventually, but the investment in
> shipping is prohibitive.
>
> Unfortunately for the builders of "one-of's," like me (spelled one-off's in some
> places, for reasons I don't understand), the programmable logic comes in
> packages that are really inconvenient for that purpose. Nowadays, you're
> required to use a 1000+ pin BGA package to get the quantity of logic that you'd
> like in a 44-pin PLCC. If you're really lucky with the fit, the device will
> allow you to use 10% of the gate count the marketing guys said you're paying
> for. The result is that you have to use a PGA package that lists for over $1k
> per part, or build an adapter board for the cheaper TQFP part, with a total
> cost, when you're done, of over $1k per each anyway, and then, to make matters
> worse, only one in 7 of these adapter boards will turn out to be properly
> soldered, so you waste 6 FPGA's costing $300 each. That's why product
> development that once required a cash outlay of $500 now costs $500 million.
>
> Dick

Don't forget the $500 minimum order per line item. I suspect you need
all the 1000+ pins since
only 50% of the pins are usable -- the power,ground and configuration
pins take up a lot of space.
Ben Franchuk.
-- 
Standard Disclaimer : 97% speculation 2% bad grammar 1% facts.
"Pre-historic Cpu's" http://www.jetnet.ab.ca/users/bfranchuk
Received on Fri Nov 02 2001 - 19:22:30 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:34:13 BST