In article <240620031100332503%pk[at]ill_check_the_newsgroup.com>, Pique wrote:
> Any idea why apple used Linux for the Dell machine v the G5?
> http://www.veritest.com/clients/reports/apple/apple_performance.pdf
> If it makes it slower won't the windows people cry foul?

Well, Linux tends to run faster on the same hardware than Windows for
one thing.

The other reason that I suspect is that SPEC is very heavily unix-based.
I'm not even sure if the SPEC benchmarks have even been ported to
Windows. By testing on Linux, it makes it easier to get the benchmarks
up and running.

--

-------------------- http://www.techhouse.org/lou ----------------------
"Dragonmaster Lou" | "Searching for a distant star, heading off to
lou at techhouse org | Iscandar, leaving all we love behind, who knows
Tech House Alum | what dangers we'll find..."
-----------------------------------------------------------------------

In article <bearclaw-EC8D03.07232624062003[at]corp.supernews.com>,
bearclaw[at]cruller.invalid wrote:

> In article <slrnbfgm6f.dlh.lou[at]techhouse.brown.edu>,
> Dragonmaster Lou <lou[at]SPAM.ME.AND.DIE.techhouse.org> wrote:
>
> > In article <240620031100332503%pk[at]ill_check_the_newsgroup.com>, Pique
> > wrote:
> > > Any idea why apple used Linux for the Dell machine v the G5?
> > > http://www.veritest.com/clients/reports/apple/apple_performance.pdf
> > > If it makes it slower won't the windows people cry foul?
> >
> > Well, Linux tends to run faster on the same hardware than Windows for
> > one thing.
> >
> > The other reason that I suspect is that SPEC is very heavily unix-based.
> > I'm not even sure if the SPEC benchmarks have even been ported to
> > Windows. By testing on Linux, it makes it easier to get the benchmarks
> > up and running.
>
> Would a better speed test (for comparison to windoze machines) be to run
> one of the Mac Windows emulation programs on the new processor and see
> how it stacks up against a fast wintel machine?

Hardly. This is a test that the Mac would be guaranteed to lose. Even
if the G5 is twice the performance of the fastest PC (as indicated by
the application benchmarks), emulation would cost so much more.
Besides, if all that matters is the performance of Windows software, a
Mac wouldn't be in the running.

--
Hank Shiffman http://www.disordered.org
Have Opinion, Will Travel hank[at]disordered.org
Mountain View, California

Dragonmaster Lou <lou[at]spam.me.and.die.techhouse.org> wrote:
> The other reason that I suspect is that SPEC is very heavily
> unix-based. I'm not even sure if the SPEC benchmarks have even been
> ported to Windows. By testing on Linux, it makes it easier to get
> the benchmarks up and running.

http://www.spec.org/cpu2000/

A search for "Windows" as the OS for a SPECcpu2000 seems to yield 660
results. This is across the four metrics - int/fp speed and int/fp
rate. SPEC goes to great pains to make sure the benchmarks are
"portable."

rick jones
--
portable adj, code that compiles under more than one compiler
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to post, OR email to raj in cup.hp.com but NOT BOTH...

In article <bearclaw-EC8D03.07232624062003[at]corp.supernews.com>,
<bearclaw[at]cruller.invalid> wrote:
>In article <slrnbfgm6f.dlh.lou[at]techhouse.brown.edu>,
> Dragonmaster Lou <lou[at]SPAM.ME.AND.DIE.techhouse.org> wrote:
>
> > In article <240620031100332503%pk[at]ill_check_the_newsgroup.com>, Pique wrote:
> > > Any idea why apple used Linux for the Dell machine v the G5?
> > > http://www.veritest.com/clients/reports/apple/apple_performance.pdf
> > > If it makes it slower won't the windows people cry foul?
> >
> > Well, Linux tends to run faster on the same hardware than Windows for
> > one thing.
> >
> > The other reason that I suspect is that SPEC is very heavily unix-based.
> > I'm not even sure if the SPEC benchmarks have even been ported to
> > Windows. By testing on Linux, it makes it easier to get the benchmarks
> > up and running.
>
>Would a better speed test (for comparison to windoze machines) be to run
>one of the Mac Windows emulation programs on the new processor and see
>how it stacks up against a fast wintel machine?

I think that if the OS X version of SoftWindows had been available, that
is precisely what SJ would have done. It wouldn't have been as fast,
true, but it might have been pretty impressive nonetheless.

It is like the description of Samuel Johnson with respect to women giving
the sermon at Sunday services: which he likened to a dog walking on its hind
legs. It is not that it is done well; one is astonished to see it done
at all.

David Derbes [loki[at]uchicago.edu]

Rick Jones <foo[at]bar.baz.invalid> wrote in news:mD%Ja.3271$yt.2326
[at]news.cpqcorp.net:

> Dragonmaster Lou <lou[at]spam.me.and.die.techhouse.org> wrote:
>> The other reason that I suspect is that SPEC is very heavily
>> unix-based. I'm not even sure if the SPEC benchmarks have even been
>> ported to Windows. By testing on Linux, it makes it easier to get
>> the benchmarks up and running.
>
> http://www.spec.org/cpu2000/
>
> A search for "Windows" as the OS for a SPECcpu2000 seems to yield 660
> results. This is across the four metrics - int/fp speed and int/fp
> rate. SPEC goes to great pains to make sure the benchmarks are
> "portable."
>
> rick jones

But... but... but...

Why didn't apple run SPEC against a similar stock G4 running the same OS?

Forgive my ignorance.

Andy

david raoul derbes <loki[at]midway.uchicago.edu> wrote:


> >Would a better speed test (for comparison to windoze machines) be to run
> >one of the Mac Windows emulation programs on the new processor and see
> >how it stacks up against a fast wintel machine?
>
> I think that if the OS X version of SoftWindows had been available, that
> is precisely what SJ would have done.

Why didn't he do it with Virtual PC?

> It wouldn't have been as fast, true, but it might have been pretty
> impressive nonetheless.

I don't think it would have been. Why do you think it would have?




--
Andrew J. Brehm
Fan of Woody Allen
PowerPC User
Supporter of Pepperoni Pizza

Buteo Lagopus <lbl[at]pbzpnfg.arg> wrote:
> Rick Jones <foo[at]bar.baz.invalid> wrote in news:mD%Ja.3271$yt.2326
> [at]news.cpqcorp.net:
>> http://www.spec.org/cpu2000/
>>
>> A search for "Windows" as the OS for a SPECcpu2000 seems to yield 660
>> results. This is across the four metrics - int/fp speed and int/fp
>> rate. SPEC goes to great pains to make sure the benchmarks are
>> "portable."

> But... but... but...
> Why didn't apple run SPEC against a similar stock G4 running the
> same OS?

No idea. All I wanted to do was show that it is possible to run
SPECcpu2000 under Windows, along with Linux, MacOS X, HP-UX, Solaris,
AIX, etc etc etc. As for why any one individual or group decides to
publish numbers for any one platform will vary with the individual or
group.

rick jones
--
Wisdom Teeth are impacted, people are affected by the effects of events.
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to post, OR email to raj in cup.hp.com but NOT BOTH...

david raoul derbes <loki[at]midway.uchicago.edu> wrote:

> In article <1fx2r44.1sg5mle13sar2oN%andrew[at]netneurotic.de>,
> Andrew J. Brehm <andrew[at]netneurotic.de> wrote:
> >david raoul derbes <loki[at]midway.uchicago.edu> wrote:
> >
> >
> >> >Would a better speed test (for comparison to windoze machines) be to run
> >> >one of the Mac Windows emulation programs on the new processor and see
> >> >how it stacks up against a fast wintel machine?
> >>
> >> I think that if the OS X version of SoftWindows had been available, that
> >> is precisely what SJ would have done.
> >
> >Why didn't he do it with Virtual PC?
> >
> >> It wouldn't have been as fast, true, but it might have been pretty
> >> impressive nonetheless.
> >
> >I don't think it would have been. Why do you think it would have?
> >
>
> I didn't see the keynote. I don't know if SJ had a working dual 2 GHz
> 970 up or not. But, let's say he had.

He did.

> Based on nothing but a wild-ass guess (:-)), I think this machine would
> run Windows ME at something like 2/3 of the speed of the average user's
> machine.

It might even run it much faster. But graphics performance (including
the desktop and windows etc.) is much much slower due to the emulation.
It wouldn't be impressive.

> That to me would be impressive. Perhaps my levels of impressionabilty
> are far below yours. :-)

I think you underestimate the difficulty of emulating video output.

--
Andrew J. Brehm
Fan of Woody Allen
PowerPC User
Supporter of Pepperoni Pizza

> Would a better speed test (for comparison to windoze machines) be to run
> one of the Mac Windows emulation programs on the new processor and see
> how it stacks up against a fast wintel machine?

What I'd like to see:
Powerful Mac with a slave Pentium and hardware/software
support for letting the Pentium run Windoze code
while Mac OS X in in charge.

(In other words, don't emulate, just time-slice to
the other processor, with traps to switch back
when they try to get to the hardware)

In article <E5qdnSSkXZAucWWjXTWJkg[at]gbronline.com>, Wesley Groleau wrote:
>
>> Would a better speed test (for comparison to windoze machines) be to run
>> one of the Mac Windows emulation programs on the new processor and see
>> how it stacks up against a fast wintel machine?
>
> What I'd like to see:
> Powerful Mac with a slave Pentium and hardware/software
> support for letting the Pentium run Windoze code
> while Mac OS X in in charge.
>
> (In other words, don't emulate, just time-slice to
> the other processor, with traps to switch back
> when they try to get to the hardware)

Apple did something like this years ago, and it didn't sell particularly
well if I recall.

--

-------------------- http://www.techhouse.org/lou ----------------------
"Dragonmaster Lou" | "Searching for a distant star, heading off to
lou at techhouse org | Iscandar, leaving all we love behind, who knows
Tech House Alum | what dangers we'll find..."
-----------------------------------------------------------------------

In article <s14Ka.33$Y4.17731[at]news.uchicago.edu>,
loki[at]midway.uchicago.edu (david raoul derbes) wrote:

> In article <1fx2r44.1sg5mle13sar2oN%andrew[at]netneurotic.de>,
> Andrew J. Brehm <andrew[at]netneurotic.de> wrote:
> >david raoul derbes <loki[at]midway.uchicago.edu> wrote:
> >
> >
> >> >Would a better speed test (for comparison to windoze machines) be to run
> >> >one of the Mac Windows emulation programs on the new processor and see
> >> >how it stacks up against a fast wintel machine?
> >>
> >> I think that if the OS X version of SoftWindows had been available, that
> >> is precisely what SJ would have done.
> >
> >Why didn't he do it with Virtual PC?
> >
> >> It wouldn't have been as fast, true, but it might have been pretty
> >> impressive nonetheless.
> >
> >I don't think it would have been. Why do you think it would have?
> >
>
> I didn't see the keynote. I don't know if SJ had a working dual 2 GHz
> 970 up or not. But, let's say he had.
>
> Based on nothing but a wild-ass guess (:-)), I think this machine would
> run Windows ME at something like 2/3 of the speed of the average user's
> machine. That to me would be impressive. Perhaps my levels of impressionabilty
> are far below yours. :-)
>
When I run VPC 6 on my 17" PB, Win XP reports itself running on a 1.33
GHz 686.... that's pretty damn good, to me.

I was demo-ing how well Win98 ran the other day when it blue
screened..... I got a good laugh out of that. :D


- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
"We really haven't done everything we could to protect our
customers ... Our products just aren't engineered for security."
-- Brian Valentine, Sr VP in charge of MS Windows' Dev Team

Dragonmaster Lou <lou[at]SPAM.ME.AND.DIE.techhouse.org> wrote in message news:<slrnbfgm6f.dlh.lou[at]techhouse.brown.edu>...
> In article <240620031100332503%pk[at]ill_check_the_newsgroup.com>, Pique wrote:
> > Any idea why apple used Linux for the Dell machine v the G5?
> > http://www.veritest.com/clients/reports/apple/apple_performance.pdf
> > If it makes it slower won't the windows people cry foul?
>
> Well, Linux tends to run faster on the same hardware than Windows for
> one thing.
>
> The other reason that I suspect is that SPEC is very heavily unix-based.
> I'm not even sure if the SPEC benchmarks have even been ported to
> Windows. By testing on Linux, it makes it easier to get the benchmarks
> up and running.

They have ported SPEC to run under Windows since the NT days. For a
little history (I'm showing my age) Intel has always been rather
disrespectful of Windows, particularly the 9x and ME versions, saying
it didn't push their hardware enough. When Microsoft went Windows
2000 a company created a special library for running benchmarks called
the MicroQuill Smartheap Lbrary. Now the exciting thing about this
library was that like all libraries it speeded up some thing and
slowed other things down but it always did great for benchmarks. And
it definitely made Windows look much faster than Linux in the
benchmark scheme of things. So it was added in everytime you wanted
to run SPEC. It cost our company $1000 (we got an educational
discount) but only helped in a couple of routines for our product so
we ditched it. But we keep it around so we can use it to show off our
products using a special benchmark, cool huh? Weird thing is, every
PC pundit accuses Apple of being disingenuous with their benchmark
claims.

What you should realize, however, is that Steve Jobs isn't the only
one saying the G5 is the fastest PC in the world. Pixar's chief
engineer, who ditched all the SGI computers in favor of dual XEON, has
publically come out and said the same thing. The bad thing is you and
I may not be able to get a G5 right away unless we get the low end
ones. Scuttle butt from all of the scientific developers at all of
the national labs is that these things are so fast for the price, they
may be already spoken for. I don't believe Apple will be able to keep
up with demand.