[Info-vax] VMS port to x86
David Froble
davef at tsoft-inc.com
Sat May 12 01:59:26 EDT 2012
JF Mezei wrote:
> Jan-Erik Soderholm wrote:
>
>> HPVM is, as far as I know, not an emulator. If HPVM would run
>> on X86, the guest OS'es would also need to be X86 versions.
>
> The emulator company that wrote Rosetta for Apple (which did on-the-fly
> endian conversion at run time for big endian OS-X PowerPC to little
> endian OS-X on x86) was purchased by... IBM .
>
>
> So HP would have to write its own IA64 emulator with endian conversion
> to support HP-UX on x86. Hopefully that emulator could also run VMS in
> little endian mode.
>
> While this is good for "maintenance mode" operating systems, it isn't
> really suited for on-going concerns because HP would still have to
> write/update compilers for IA64 and not be able to benefit from fine
> tuning on x86, and folks like Oracle would still have to write for
> Itanium platform, something they don't want to do because Itanium is dead.
>
>
> However, the plan is to allow IA64 and X86 blades to co-exist in the
> same chassis. So your legacy apps can run on an IA64 blade while your
> port/develop them on x86 blades and once ready, you just switch them to
> the new environment in the same cabinet.
>
> I suspect Odyssey will have hardware based partitioning to allow
> different architectures and totally separate instances to co-exist
> without needing HPVM.
Why not just have everybody sue Intel out of business ?
Surely their plan to do away with competitive CPUs has caused harm.
Perhaps my AMD stock would finally pay off ....
More information about the Info-vax
mailing list