>Yet in research labs (a lot of clustered solutions)
>and automotive fields, the Linux kernel has a
dominant
>role (not necessarily the RH distro either).

The processor count in clustered systems is
irrelevant;
clustering is not affected by the single CPU limit.

>I think the arguement is still *when** Sun will have
>support up to 128 processors for the IA64/AMD64
>platforms (if not also Power/PowerPC) - even if it is
>in ALPHA/BETA form for OEM/IHVs.

Unlikely for the IA64 platform; soon for the AMD64
platform; for us, however, there is generally one
precondition: we need to have the platform with that
many CPUs to test.

>This also isn't just looking at Linux, but AIX and
>HP-UX. Hardware specs like:
>
>Processors: 128
>Memory: 1 TB RAM
>Storage: 512TB storage

Well, were is that AMD64 hardware which such SPECs?
E25K: 72CPUs * 2 cores.

So the basic OS supports that many CPUs and more;
we can't very well support systems which do not exist.

Casper
-----------------------

Casper,

You hit the nail on the head in that if people don't
have ACCESS to these systems - but not if the systems
do not exist though. Don't forget the HP Superdome and
IBM p5 595, a several "Skunk Works" projects in which
investors have gone well up to 128 physical cpu
configurations (and more).

Yet, I know the people on this email are champions in
their field and have experience from Wasabi Systems
and the early NetBSD/FreeBSD integrations of SMP. But,
that was about 4-5 years ago. If Sun gave Solaris
using AMD64/IA64 hardware with up to 32 cpus (32/64
cpus with dual core), that would be great. Maybe, even
make a user tunable selection where a user can specify
the number of physical CPUs >8 cpus might be an
option.

You already know that things don't have to exist in
native hardware to exist in software / custom
hardware....

~ Ken





__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to