Re: SF16

[msimon@xxxxxxxxxx (M. Simon)]

>>>From: Christopher Biggs <chris@stallion.oz.au>
>>>To: <MISC@pisa.rockefeller.edu>
>>>Sent: Wednesday, April 07, 1999 7:07 PM
>>>Subject: Re: SF16
>>>
>>>dirnfir <dirnfir@softhome.net> moved upon the face of the 'Net and spake
>>>thusly:
>>>
>>>>
>>>>Why not 32 bit?  And lots (several MB) of memory... you could make it fit
>>>>ATX form factor even...
>>>
>>>That's too big.  My ideal computer is the size of a deck of cards.
>>>Something about the size of a current PC104 card is tolerable.  As big
>>>as a desktop motherboard is Too Big.

There would be a number of 'final' versions.

The first one would be for prototyping purposes.

>>>(IMHO, of course)
>>
>>You can get mass-produced cases that small?  If you can that might be a good
>>idea.  Since this thread is rapidly mutating into "MISC Personal Computers",
>>I'll toss out some ideas...
>>
>>- I am assuming that it would use an FPGA...  How fast can those things go?

You can clock parts of the circuit (say a video pump) at
100MHz.

A processor might do 20 to 50 MHz. 


>>- Use PS/2 mouse/keyboard ports...  so I can avoid using serial mice...

They are easy. Two or three pins. An I2C type protocol.

And that is the plan.

>>- For the far future: multi-cored processors anyone?
>>
>>- Use standard SDRAM...  Nowadays it comes in an 8ns variety.  That's 125
>>MHz with a CAS latency of 2.  64-bit though.

I don't care to build a refresher/memory manager at this time.

SRAM is for dummies.

Second version though.

And there would be enough spare I/O to try it out. Lots of hours though.

Just getting up the stuff I mentioned will be a lot for the first cut.

But we can build on our successes.

Simon
Simon - http://www.tefbbs.com/spacetime/index.htm