> If I understand this right.  The Itanium 64 bit chip required you to buy
>new hardware and rewrite software.

Yes and no. Some of the more recent HP servers were made that you
could buy as PA-RISC now, and swap to Itanium later. It had to do with
I/O backplanes, etc.

Further, the Itanium chips had an emulation mode which could run
native PA-RISC and native x86 instructions without recompiling.
However, there was a non-trivial performance hit for that.

> The Xeon 64 bit chip will run both 32 bit code and 64 bit code.

Yes, but will there be a performance penalty for one or the other?


> The Xeon 64 bit was designed to compete with the AMD 64 bit similar product that runs 32 and 64 bit code.
>
>IF (Big "IF") HP were to develop hardware based on the Xeon chip instead of
>the expensive Itanium chip it could continue the HP3000 without
rewriting the OS.

I think that's too big a logical jump. MPE would have to be recompiled
(just to execute on an x64-64 chip which knows nothing about PA-RISC
instructions) and rewritten, since the I/O busses in modern Intel
hardware is somewhat different than what MPE really supported at the
end of it's run. Recompiling MPE to run on an x86-64 intstruction set
would be extremely non-trivial. It's also why I don't expect an x86
version of HP-UX to ever come out. HP is still committed to PA-RISC
for the near-to-medium future, and committed to the Itanium line for
servers byond that. As for workstations and desktops, HP has an
interesting problem. If they want to keep the workstation and server
version of HP-UX identical, and they aren't doing Itanium on desktops
and PA-RISC is going away someday, they have the following choices:
1) Push workstation customers to Linux. This is OK, but many
application vendors don't support Linux .. yet.
2) Push workstation customers to WIndows. Some customers just plain
won't do that. They will select another hardware / OS vendor in that
case.
3) Recommit to Itanium on the desktop. (I predict this will happen, someday).
4) Continue to support HP-UX on Itanium for servers, but only on
PA-RISC for workstations. This would involve splitting the HP-UX
product line into server vs. workstation releases, but HP spent a lot
of effort in the HP-UX 10 timeframe doing exactly the opposite .. one
HP-UX release for both types of hardware.
5) Ressurect the PA-RISC line. I think this is not going to happen.
They've already divested a lot of their PA-RISC resources and it would
be extremely hard to get it going again.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *