> At this stage, I'd suggest the software CPU support is your smallest
> problem.  There might be a small amount of pmax-specific rework, but
> in general I don't think this bit will be too much at all.  Put it
> this way - if you get a working design, we'll make the software work.

That's the way I look at it ;-)

> Emphasis on _little_ here.  DEC and everyone else seem to have different
> ideas about CPU speeds for the R4x00.  These CPUs are internally
> clock-doubled - what DEC calls a 60MHz part everyone else would seem
> to call a 120MHz part.  So a 133MHz CPU would be a 5000/266 (or
> something!).

The R4400 could run either with a straight 60Mhz pumped to it and a pin was
set to 1x, or you would pump 30Mhz in and set a pin to 2x which activate a
PLL, in the CPU, that did an internal clock doubling (like the 486DX2).  If
I manage to graft a new chip on the DB, I will have to change the clock
(30Mhz) or else you'll see little boost in speed since you'll still be
running 60Mhz (30x2) at the core.

So I guess it would be a 5000/2133, or something.

> Best of luck with your project!

Thanks, I need it!  Info is hard to come by on the DECstation and R3000 PGA
chips.

Lots of good leads though.

john