[Info-vax] Poulson info from Dave Cantor

[Michael Kraemer]

John Wallace schrieb:
> On Nov 18, 9:56 am, Michael Kraemer <M.Krae... at gsi.de> wrote:

> Got any other evidence that your vision of cost sharing to preserve
> IA64 has any chance of happening?

It actually has already happened for nearly a decade.
At least neither intel nor HP seem
to groan under the burden,
whereas DEC had to get rid of their chip after only five years.

> As for fabrication costs: when your volumes are relatively small, your
> per-chip fabrication costs aren't that relevant, what's relevant is
> the per-chip share of the non-recurring development costs. Bear in
> mind also that at any given time IA64 has tended to be one or two
> process generations behind the current mainstream Intel process. So
> even if producing IA64s on a fab configured for x86-64 was technically
> possible (is it?), the saving isn't going to massively affect the
> total cost per chip.

Iirc intel did not have to build and maintain
an extra fab for the production
of just a few ten thousand chips per year.

> "I just tried an ancient "edt" built in 1991 for POWER1, still runs on
> POWER7, unmodified. "
> 
> And based on that test we are expected to be convinced that the modern
> system will correctly run arbitrarily chosen examples of other ancient
> real-world code ? I think not. Proper engineering "qualification"
> tests would need a little more thought (and investment).

That edt example is of course just anecdotal evidence,
that binary is one of the oldest I have.
But I found that in general to be true for the POWER platform
in the past two decades. And the IBM compiler docs explicitly state
that as long one sticks to -qarch=com (which is the default):
      com     Produce an object that contains instructions
              that will run on all the POWER and PowerPC
              hardware platforms.
I have no reason not to trust it.