The TI 34010

From: Derek Peschel <dpeschel_at_u.washington.edu>
Date: Sat Mar 6 19:48:51 1999

Tony Duell wrote:

[transputers generally rely on a host]

> (1) is not so much of a problem. The reason most transputer systems have
> a 'host' is that you can boot a transputer over one of its links without
> having any user-written boot code on the transputer. This means it's easy
> to develop for (no possibility of your (firmware) bootstrap having bugs
> in it), etc. But you can equally boot a transputer from external ROM in
> its own memory space if you want to. And hook a terminal up over an RS232
> port (the transputer's bus is similar to just about every other bus
> you're likely to see, so you can have RS232 chips and SCSI chips, and
> whatever, mapped into the transputer's address space).

So graphics chips are a good possibility. *drool* RS232 is fine but it's
not very flashy. :)

> In fact one of the first INMOS transputer boards - the B001 - had a
> transputer + RAM + RS232 port + ROM on it. It didn't expect a transputer
> link to the host.

This may be a good time to ask if there's an exhaustive list of part numbers
and specifications for the various INMOS chips? I've seen lists (in a
couple of mass-market books about the transputer) but I suspect they're not
complete. I've never seen the actual data books.

> Problem (3) is more serious - the fact that most (all?) of the transputer
> chips are no longer in production.

I guess that IS a problem. Do they show up on the surplus market?

> > them; Sketchpad relied on them and Evans & Sutherland continued the trend.
> > (I'd love to play with some of their machines!)
>
> Does anyone know anything about the PS/390 display from E&S?

Not me. There's some sort of E&S display at the Computer Museum History
Center but I don't know the model number.

-- Derek
Received on Sat Mar 06 1999 - 19:48:51 GMT