[ale] Re: was -- Multi-drop PPPoE.... -- now 386-486 sx/dx - and more OT

[Frank Zamenski]

> Frank Zamenski wrote:
>
> > Some of those 386DX mobos also had an additional socket for the 387
mathco.
> > Did/does having one of those onboard eliminate the need to compile
kernel
> > FPU
> > emulation? I don't know myself.
>
> Right. If you have a 386 with the companion FPU
> chip, you can use a kernel without FPU emulation. I expect few
> such systems actually have the FPU installed, tho.
>
> I believe the 386 and 486 SX will both accept either the
> 80287 or 80387 FPU chips. I also remember hearing

Um, not sure about that, I think the 286 topped out at 10mhz,
whereas the 386 began at 16. But, also not sure if the 287 and
387 sockets were true 1:1 busses either... memory fades...

> that the chips marketed as 486SX were actually 486DX
> chips that had the on-chip FPU disabled, and that chips
> marketed as 80487 mathcos were actually fully-functional
> 486DX chips.
>
> -- Joe

As I recall (subject to the same fading memory), the Intel marketing
machine wanted everyone to buy into the 486 architecture quickly,
and developed the 486DX first, but they sold it at a premium mostly
to businesses. Not satisfied with sales figures, they indeed created
a cheaper entry-level chip by emasculating perfectly good 486DX
FPUs into the 'SX', before they tooled up to make real and cheaper
SXs. Then some genious figured out a profitable Rube Goldberg
upgrade path with the 'overdrive' socket (the pre-Pentium O/D),
where one could buy a good 486DX -- of course also marketed as
a 487 -- to assist the disabled one. Then DX caught on, volume sales
made those cheaper, and the 'Pentium O/D' was born... yadda yadda...
gotta hand it to 'em!

The Wintel conspiracy in the making... all to run the new Win 3 and it's
successive 3.1, 3.11, 3.11/Workgroups variants. Hmm... has the
pattern really changed since? ;).

Back on topic now...

--fgz

>
> *** Joseph Knapka ***
> In any formula, constants (especially those obtained from handbooks)
> are to be treated as variables.
>

--
To unsubscribe: mail majordomo at ale.org with "unsubscribe ale" in message body.