Re: Chuck's 486
On Mon, 29 May 1995, Jeff Fox wrote:
> One of the things that slow Chuck down is the speed of simulation.
> It takes about 5 minutes to simulate 12 instructions, so testing
> is very slow. With three coprocessors all capable of interrupting
> the CPU there are many conditions that need to be simulated to
> verify the design.
On Tue, 30 May 1995, Jeff Fox wrote:
> Chuck is still using the OKAD on the 486.
On Wed, 31 May 1995, Eddie Matejowsky wrote:
> Is he's really running a 486 or a 586 running 486 code? If it's really a 486
> someone should lend him a pentium (or two) to speed things up. I'd lend him
> mine if I lived closer.
I have an idea.
A NexGen Nx586 P90 with 256K 12ns SRAM can be had (tax,S&H included) for
$545. Since it has a 64-bit path to DRAM, 2 72-pin SIMMS are need. 2 x
1Mx32 70ns can be had for another $315, which makes a total of $850.
If Jeff agrees to sell coupons for future F21s from the first lot for
$10 each (instead of the $20-$25 MuP21 goes for now), he needs to sell 85
chips now to get Chuck the Nx586 P90.
There are ~150 subscribers to the list currently. I personally am ready
to buy 3 F21s @ $10 now and get them when they are ready.
If Jeff thinks, that the simulations are the bottleneck, and that Chuck's
time can be utilized better from doing work, than getting frustrated
watching signals crawl, he might announce the offer.
If he gets email orders for 85 chips, he can say "Go ahead, send the
checks," and in a week Chuck will have the board. If not -- may be we
don't need F21s that much.
--
Penio Penev <Penev@venezia.Rockefeller.edu> 1-212-327-7423
