Why so few garage processors? (was Re: Squeak History / Tiny
Machines)
Hannes Hirzel
hannes.hirzel.squeaklist at bluewin.ch
Wed Mar 19 17:03:12 UTC 2003
Ned Konz <ned at bike-nomad.com> wrote:
> On Tuesday 18 March 2003 02:24 pm, Alan Kay wrote:
> > At 6:40 PM -0300 3/18/03, Jecel Assumpcao Jr wrote:
> > >I have been learning a lot trying to fit Smalltalk into 15
> > > thousand gates (http://www.merlintec.com:8080/Hardware/Oliver)
> > > and feel that my larger projects will be far better because of
> > > this.
> >
> > Well, IIRC, the first ARM that Tim did so many neat things with was
> > only about 25,000 transistors (and the Alto was a lot less than
> > that).
>
> It used to be that you had to work for a big company to be able to
> design high-performance processors and systems.
>
> But today, with the average $1000 computer having considerably more
> power than the CAD workstations of just a few years ago, we can do
> these designs ourselves.
>
> I wonder why we don't see more innovative processor/system solutions
> coming from individuals now that the (financial) cost of entry for
> making fast, capable systems is almost $0.
>
> We can download free FPGA/CPLD design tools capable of dealing with
> large devices, and the devices themselves are pretty cheap.
>
> We can get free PC board design software.
>
> We can get prototype PC boards made for around $33 each.
>
> For instance, using the free ISE Webpack from Xilinx (Altera and
> Lattice also have free software like this) you can program a device
> like the Virtex XCV300E, which has:
>
> 32x48 CLB array
> 6912 logic cells
> 411944 system gates
> 131072 max. block RAM bits
> 98304 max. distributed RAM bits
>
> and is available in non-BGA forms (i.e. ones that can be
> hand-soldered). You can buy them in single quanties for less than
> $250.
>
> Take one of these, connect it to some fast DRAM and I/O, and off you
> go...
>
> I know Jecel is doing interesting work; why don't we see more
> Smalltalk machines coming out of peoples' garages?
Probably because of lack of system integration know-how.
You have to be a hardware and a software type to come up with this.
There are not many of these people nowadays.
What you write is actually exciting!
We would need a ready made board a CD with (free) software and HOWTOs.
So people could concentrate on the SW side.
Actually this would make up a good project for a student.
Perhaps (heavy) help from some members of
this list would be needed.
I think this would be an incredible CS learning experience.
-- Hannes
More information about the Squeak-dev
mailing list
|