In article <7w8yd6ymuc.fsf@sic.twinsun.com>, Paul Eggert wrote:
>It is nice to see Sun's uniprocessor SPEC results finally looking
>competitive again. For example, for the W1100z Sun reports CINT2000
>base = 1434, CINT2000 Rate base = 16.6. Both of thes numbers are
>quite respectable compared to the competition (though they're not the
>absolute top), and they blow away the best uniprocessor SPARC numbers
>that Sun has published (642 and 7.44, respectively, for a 1.2 GHz Sun
>Blade 2000).

And just about beat the best Fujitsu SPARC numbers too.

>Too bad specbench.org isn't publishing the floating
>point numbers yet.

There are on the Sun site (a recent post mentioned the links).

>I noticed a few other things:

>Solaris x86 64-bit support is not available until Solaris 10 ("early
>2005" -- could that mean as late as June 2005?).

I'd have thought that early meant first quarter.

....

>The W2100z uses Ultra320 SCSI. Is desktop FC-AL dying?

It's been dead for a long time. It was a stupid idea in the first
place. I'd have thought that desktop SCSI is also dead, why bother
with expensive SCSI disks when you can have larger, cheaper S-ATA
disks?

--
Phillip Fayers School of Physics & Astronomy, Cardiff University (*).
[ (*) - the official trading name of the University of Wales, Cardiff. ]
P.Fayers@astro.cf.ac.uk http://www.astro.cf.ac.uk/pub/Phillip.Fayers/
Tel: +44 (0)29 2087 5282 Attribute these comments to me, not UWC.

Richard.L.Hamilton@mindwarp.smart.net (Richard L. Hamilton) writes:

> You may be forgetting the early Merced port they'd already done.
>
> I doubt that 64-bitness of x86 source is going to be the problem as
> much as the low-level stuff and driver mods.

It's a soluble problem for sure, but each port has its own
wrinkles. For example I port something from windows to solaris, so I
use an ifdef like ACMEUNIX, then I port from there to linux, well the
fact of it being UNIX is no longer good enough - I need to know
big-endian or little-endian also. Ok, so that's my mistake, but it's
not uncommon.

Chris
--
Chris Morgan
"Post posting of policy changes by the boss will result in
real rule revisions that are irreversible"

- anonymous correspondent

In article <7w8yd6ymuc.fsf@sic.twinsun.com>,
Paul Eggert <eggert@twinsun.com> wrote:
>Solaris x86 64-bit support is not available until Solaris 10 ("early
>2005" -- could that mean as late as June 2005?).

Management previously stated Solaris 10 would go FCS in December but
heard January at the Solaris BOF at Usenix Boston.

John
groenveld@acm.org

"UNIX admin" <tripivceta@hotmail.com> writes in alt.solaris.x86:
|
|> <URL:http://www.theregister.co.uk/2004/07/26/sun_opteron_finally/>
|> | At long last, Sun's four processor Sun Fire V40z server and W1100z and
|> | W2100z workstations have gone on sale. All of the systems run on AMD's
|> | latest Opteron chip and complement Sun's existing two processor Sun
|> | Fire V20z server. With four Opteron systems now in its product line,
|> | Sun has officially become the biggest Tier I backer of AMD's x86-64bit
|> | product.
|
|What does `isainfo -vk` return on those systems?

% isainfo -kv
32-bit i386 kernel modules

(Of course, this is on a W1100z prototype running S9 9/04, so that's
what you'ld expect since the 64-bit x86 kernel isn't coming until S10.
I don't have a machine running that handy to check on.)

|And how does the naming scheme ("V1100z" etc.) work?

V20z, V40z, W1100z, W2100z

I think it's V for "volume server" (aka low-end server sold in volume
instead of higher-end MP machines that are much rarer) vs. W for
"Workstation". Like the SunBlade 1500/2500, the first digit is the
number of processors. The "z" suffix indicates Opteron CPU. But I
don't have a written reference to confirm that, and my mind has been
reduced to a pile of rubble lately, so I may be misremembering.

--
__________________________________________________ ______________________
Alan Coopersmith * alanc@alum.calberkeley.org * Alan.Coopersmith@Sun.COM
http://www.csua.berkeley.edu/~alanc/ * http://blogs.sun.com/alanc/
Working for, but definitely not speaking for, Sun Microsystems, Inc.

> % isainfo -kv
> 32-bit i386 kernel modules
>
> (Of course, this is on a W1100z prototype running S9 9/04, so that's
> what you'ld expect since the 64-bit x86 kernel isn't coming until S10.
> I don't have a machine running that handy to check on.)

The word 'round the campfire is that they've got the Solaris10 64-bit kernel
booting on the Opteron CPU. Do you know if they'll be getting the Solaris9
kernel to work in 64-bit mode as well?

> V20z, V40z, W1100z, W2100z
>
> I think it's V for "volume server" (aka low-end server sold in volume
> instead of higher-end MP machines that are much rarer) vs. W for
> "Workstation". Like the SunBlade 1500/2500, the first digit is the
> number of processors. The "z" suffix indicates Opteron CPU. But I
> don't have a written reference to confirm that, and my mind has been
> reduced to a pile of rubble lately, so I may be misremembering.

Thanks, even that clears it up a lot!

"UNIX admin" <tripivceta@hotmail.com> writes in comp.sys.sun.hardware:
|> % isainfo -kv
|> 32-bit i386 kernel modules
|>
|> (Of course, this is on a W1100z prototype running S9 9/04, so that's
|> what you'ld expect since the 64-bit x86 kernel isn't coming until S10.
|> I don't have a machine running that handy to check on.)
|
|The word 'round the campfire is that they've got the Solaris10 64-bit kernel
|booting on the Opteron CPU.

"They" do, but the W1100z machine I have access to is being used for
testing the Java Desktop System for Solaris 9 x86 for the new
workstations, so it's running S9 9/04 for now. After that's done it
will probably be upgraded to S10 for 64-bit x86 testing.

|Do you know if they'll be getting the Solaris9
|kernel to work in 64-bit mode as well?

Last I heard there were no plans to backport to S9 - but why wouldn't
you just want to run S10 anyway?

--
__________________________________________________ ______________________
Alan Coopersmith * alanc@alum.calberkeley.org * Alan.Coopersmith@Sun.COM
http://www.csua.berkeley.edu/~alanc/ * http://blogs.sun.com/alanc/
Working for, but definitely not speaking for, Sun Microsystems, Inc.

On Wed, 28 Jul 2004 06:46:07 +0000 (UTC) Alan Coopersmith <alanc@alum.calberkeley.org> wrote:
> "UNIX admin" <tripivceta@hotmail.com> writes in comp.sys.sun.hardware:
> |Do you know if they'll be getting the Solaris9
> |kernel to work in 64-bit mode as well?
>
> Last I heard there were no plans to backport to S9 - but why wouldn't
> you just want to run S10 anyway?

Alan! This isn't alt.linux. :-) One might want to run a 32-bit app on a
64-bit kernel and stick with S9. Even with PAE, there's one very good
reason: your app gets more memory because the kernel isn't eating part
of your address space. Or would that still be the case?

I doubt there's significant need for this, because so few folks are
running S9/x86 right now, but ... just saying.

/fc

> Last I heard there were no plans to backport to S9 - but why wouldn't
> you just want to run S10 anyway?

I would, but not while it's still beta. When it goes past the FCS release,
that's when I'm jumping on board.

On Wed, 28 Jul 2004, UNIX admin wrote:

> I would, but not while it's still beta. When it goes past the FCS release,
> that's when I'm jumping on board.

S10 Beta is perfectly fine for every day use. But I agree that
keeping it off critical production servers until FCS probably
isn't a bad idea.

--
Rich Teer, SCNA, SCSA

President,
Rite Online Inc.

Voice: +1 (250) 979-1638
URL: http://www.rite-online.net