In article <1gxlvmw.m9smkqz2ngqoN%neillmassello@earthlink.net >,
neillmassello@earthlink.net (Neill Massello) wrote:

> > The rumor is they'll start at the low end machines and move it up, but
> > what if they do it the other way around, and the first Power Mac with
> > an Intel processor has a G5 _and_ a dual-core Itanium or something?
>
> That makes more sense than the plan described in the Cnet article. But I
> don't see why the whole transition to Intel makes more sense for Apple
> than sticking with PowerPC. How will Apple explain away the charts at
> <http://www.apple.com/powermac/performance/>? Can anybody make a case,
> on the basis of speed, cost, and power/heat, for this change?

Here are two interesting links:
http://www.intel.com/products/proces...ium2/index.htm
http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx

It is a move towards the Itanium, which is big endian, just as the IBM
PowerPC chips, developed by Intel and HP together. So it seems to be both
right business and development moves. One should get even more computing
power.

--
Hans Aberg

In article <haberg-0406051358440001@c83-250-193-72.bredband.comhem.se>,
haberg@math.su.se (Hans Aberg) wrote:

> In article <1gxlvmw.m9smkqz2ngqoN%neillmassello@earthlink.net >,
> neillmassello@earthlink.net (Neill Massello) wrote:
>
> > > The rumor is they'll start at the low end machines and move it up, but
> > > what if they do it the other way around, and the first Power Mac with
> > > an Intel processor has a G5 _and_ a dual-core Itanium or something?
> >
> > That makes more sense than the plan described in the Cnet article. But I
> > don't see why the whole transition to Intel makes more sense for Apple
> > than sticking with PowerPC. How will Apple explain away the charts at
> > <http://www.apple.com/powermac/performance/>? Can anybody make a case,
> > on the basis of speed, cost, and power/heat, for this change?
>
> Here are two interesting links:
> http://www.intel.com/products/proces...ium2/index.htm
> http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx
>
> It is a move towards the Itanium, which is big endian, just as the IBM
> PowerPC chips, developed by Intel and HP together. So it seems to be both
> right business and development moves. One should get even more computing
> power.

Having worked with HP Itanimum boxes, the Itanimum chip is not all that
fast compared to other chip offerings.

The Itanium chip runs hot really really hot. We had to move all of our
Itanium boxes from the normal lab into a lab in a different building,
and even there we had to upgrade the lab airconditioning systems, and
power capacity (granted we had over a 100 of 2, 4, and 8 CPU HP Itanium
boxes, but we had 100's of Alpha's in the the other labs that did not
need the same power and cooling demands.

Based on my HP experiences, I would find it hard to imaging getting an
Itanium into a laptop.

So either this is not true, or intel got Steve to drink the coolaid :-)

Just my opinion.

Bob Harris

Shawn Hearn wrote:
>
> By going with Intel, Apple will hopefully get a faster 64-bit chip for
> its desktop Macs and a 64-bit chip for use in future Powerbooks.

Isn't the rule of thumb that the data throughput of
a PowerPC is equal to that of an Intel chip double
the clock rate? However, Intel's material science
(semi process) would trump on the power front, that
is important for laptops (and in five years, desk-
tops will be an antiquated concept).

For performance, especially from big-endian code,
and overall power/data throughput, why not MIPS?

My prediction: this time next year, Intel porting
experiments with Tiger will look disappointing and
IBM will come out with a Cell processor offering
a migration path for G5 desktops and G4 laptops.

gld

In article <nospam.News.Bob-454577.11240504062005@news.verizon.net>, Bob
Harris <nospam.News.Bob@remove.Smith-Harris.us> wrote:

> Having worked with HP Itanimum boxes, the Itanimum chip is not all that
> fast compared to other chip offerings.
>
> The Itanium chip runs hot really really hot. We had to move all of our
> Itanium boxes from the normal lab into a lab in a different building,
> and even there we had to upgrade the lab airconditioning systems, and
> power capacity (granted we had over a 100 of 2, 4, and 8 CPU HP Itanium
> boxes, but we had 100's of Alpha's in the the other labs that did not
> need the same power and cooling demands.
>
> Based on my HP experiences, I would find it hard to imaging getting an
> Itanium into a laptop.
>
> So either this is not true, or intel got Steve to drink the coolaid :-)

Interesting facts. The energy consumption of a chip, though, is
proportional to the square of the CPU clock frequency. So one might play
around with that, and other developing facts, to get new, less, energy
consuming processor chips.

And it will probably quite difficult to switch from a big endian CPU to a
little endian one, as a great deal of memory data would have to be
converted for upwards compatibility. So it should be a big endian Intel
chip.

And if Steve Jobs does not have something more advance to come up with
than the G5, it's probably time to look for a new job.

So these facts together might indicate that there some kind of new chip on
the move.

--
Hans Aberg

Those remembering the 1994 68x to PowerPC move, will recall that Apple had
an ad campaign "Full Speed Ahead". Then it turned out that the PowerPC
emulation of 68x ate a lot of power, so it took some time for this
promised speed to reach the consumers. The "Full Speed Ahead" was
mentioned as a joke. Apple would not want to repeat that mistake. So a
move to Intel processors would require a work-around of that problem. Also
this speaks for a new, powerful chip.

Also, business analysts have been sceptical about this rumored Apple-Intel
move, as the 1994 move cost Apple market shares. One could turn this
around: as Apple made that move in the past, they should know what it
takes to do it again, cost and benefits. So this speaks for such a move.
In the past, Apple has always settled for new technologies, when needed.

--
Hans Aberg

http://stream.apple.akadns.net/

He starts explaining at about 22 minutes in, for anyone without time to
watch the whole thing.
GW

Hans Aberg <haberg@math.su.se> wrote:

> The energy consumption of a chip, though, is
> proportional to the square of the CPU clock frequency.

Taken at face value, this is not true: energy consumption of a normal
electronic circuit is proportional to the clock frequency, not its
square. Every clock tick electrons flow into some of the little
capacitors on the chip, and some flow out of other capacitors. Double
the pace of the clock, and the number of electrons that is pumped around
will double. Half the pace of the clock, and half the original number of
electrons will be pumped around.

It /is/ true that at a lower clock frequency it is usually possible to
lower the supply voltage of the CPU a bit, which gives you an extra
energy saving, but the overall result is not a nice clean
square-of-frequency relationship.

Hans Aberg wrote:
>
> Here are two interesting links:
> http://www.intel.com/products/proces...ium2/index.htm
> http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx
>
> It is a move towards the Itanium, which is big endian, just as the IBM
> PowerPC chips, developed by Intel and HP together. So it seems to be both
> right business and development moves. One should get even more computing
> power.
>

Itanium is crap - and even intel knows it - they have basically stopepd
marketing it

Hans Aberg wrote:
> Those remembering the 1994 68x to PowerPC move, will recall that Apple had
> an ad campaign "Full Speed Ahead". Then it turned out that the PowerPC
> emulation of 68x ate a lot of power, so it took some time for this
> promised speed to reach the consumers. The "Full Speed Ahead" was

uh, not quite. the articles/literature at the time stated 680x0
system/app code would run at the same speed (roughly) on the initial ppc
hardware via emulation with only high-use core components converting to
ppc-native first. as more components went from emulation to native,
overall OS speed would increase. the left over bits that were
infrquently called or used would never be converted; saved until a time
to be rewritten (as in transition from os7 to os8, os8 to os9).

of course, given the typical os updates also increase hardware
requirements i would actually bet the general performance between
os7/os8 were equivalent: more pieces in 8 were native and faster than
their 7 counterparts.. but had more 'functionality' and were 'slower'
than if pieces had been simply converted.

-r

In article <aJGdnXKb6-ouxz_fRVn-rA@comcast.com>,
"Fetch, Rover, Fetch" <Fetch-Rover-Fetch@K9University.edu> wrote:

> Hans Aberg wrote:
> >
> > Here are two interesting links:
> > http://www.intel.com/products/proces...ium2/index.htm
> > http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx
> >
> > It is a move towards the Itanium, which is big endian, just as the IBM
> > PowerPC chips, developed by Intel and HP together. So it seems to be both
> > right business and development moves. One should get even more computing
> > power.
> >
>
> Itanium is crap - and even intel knows it - they have basically stopepd
> marketing it

x86 is crap too. It doesn't make sense that Apple would want to switch
to a processor architecture burdened by such an old instruction set.
This rumor makes me wonder if Intel is working on a new CPU. For Apple,
Intel would be a chance to free themselves from chipmakers that don't
care much about desktop class components. For Intel, Apple would be a
proof of concept so they can market the new chips to Microsoft.

Of course the rumor could be totally blown out of proportion. It's
possible that Apple is doing nothing more than offering OS X Server as a
OEM option for x86 servers. OS X Server doesn't usually run desktop
apps so Apple would only have to re-compile the OS and the bundled
open-source packages to have a completely functional system.

[Hans Aberg wrote in comp.sys.mac]
> http://www.intel.com/products/proces...ium2/index.htm
> http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx
> It is a move towards the Itanium, which is big endian, just as the IBM
> PowerPC chips, developed by Intel and HP together.

How surprised is everyone that Apple have not picked AMD - I've been wary
of AMD kit for years, but I'm really impressed with the 64-bit Opteron
offerints. Very impressed.

Perhaps it comes down to endian conversion - ia64 is big endian, opteron
is little endian like other i86-a-likes.




--
http://fotoserve.com/ -
Superb hand-checked AGFA prints, bright sharp posters,
strong block canvas prints, unique picture bags and gifts
..... from your own digital images.

Andy Davidson wrote:

> How surprised is everyone that Apple have not picked AMD - I've been wary
> of AMD kit for years, but I'm really impressed with the 64-bit Opteron
> offerints. Very impressed.
>
> Perhaps it comes down to endian conversion - ia64 is big endian, opteron
> is little endian like other i86-a-likes.

M.O.N.E.Y.

Intel's cooperative advertising program crushes AMD. Not only does
Apple get MHz parity with PCs by moving to Intel, when you start seeing
Apple advertising with "Intel Inside", and you almost certainly will,
Apple gets a nice little kickback for the plug. And it's all quite
legal. Intel is a what, $36 billion company? AMD, maybe $8 billion?

Intel has the cash to burn and as evidenced by this link (courtesy of
Davoud), a CEO with a personal concern.

<http://online.wsj.com/public/article...520-CB7pf4gh1Z
aQ3oF44a0sjw8dJXY_20060524,00.html?mod=tff_main_tf f_top>

"Pressed about security by Mr. Mossberg, Mr. Otellini had a startling
confession: He spends an hour a weekend removing spyware from his
daughter's computer. And when further pressed about whether a
mainstream computer user in search of immediate safety from security
woes ought to buy Apple Computer Inc.'s Macintosh instead of a Wintel
PC, he said, "If you want to fix it tomorrow, maybe you should buy
something else."


It's almost as if Apple planted a mole at Intel!

--
-John Steinberg
email: not@thistime.invalid

-= I link therefore I'm spammed =-

In article <slrndabp06.t2v.$andy$@kinda.undone.org.uk>,
Andy Davidson <$andy$@nosignal.org> wrote:

> How surprised is everyone that Apple have not picked AMD

I did slightly wonder, but....

Sticking to purely technical reasons instead of business and
market-related ones (such as that I doubt Intel would have agreed to
release an OS X compiler if there were no Intel chips in any OS X
machines)...

I've got AMD chips in several (indeed most) of my "x86" boxes, and I
like the cost/performance of them fine. But one thing I definitely do
not like.... those suckers run really hot. Hot enough that I have
personally seen motherboards charred because of simple fan failures.
I've seen 2 separate cases of AMD CPUs that self-destructed after fan
failures (fortunately in the second case the damage was restricted to
just the CPU chip). I'm tempted to adopt a preventative maintenance
policy of replacing the darned fans every 6 months or so even if they
look to be going fine... except that there are problems with that also.

I won't leave my home AMD systems running when I leave the house (or go
to sleep). I consider doing so to be too much of a fire hazard. Perhaps
I'm overly cautious because I have personally seen computer systems
literally burn - not any AMD systems, but I have seen it in other
systems. About a decade ago, one of the largish systems (not a desktop)
here at work caught fire (internal board short in a board that drew so
much power normally that the extra current from the short didn't blow
the breaker). That one emptied the whole building and the smoke could be
clearly seen coming out of the window. I recall standing next to the
hardware tech for the system and watching him turn white as he saw that
this wasn't just a false alarm or drill, but that his system was
responsible for emptying the building (not to speak of self-destructing).

Considering that one of the reasons at least alluded to for the switch
was difficulties in getting G5 power consumption (and heat generation)
down to an acceptable level for a laptop... would AMD be that big an
improvement?

--
Richard Maine | Good judgment comes from experience;
email: my first.last at org.domain | experience comes from bad judgment.
org: nasa, domain: gov | -- Mark Twain

In article <mcmurtri-EED143.19365304062005@corp-radius.supernews.com>,
Kevin McMurtrie <mcmurtri@dslextreme.com> wrote:

> In article <aJGdnXKb6-ouxz_fRVn-rA@comcast.com>,
> "Fetch, Rover, Fetch" <Fetch-Rover-Fetch@K9University.edu> wrote:
>
> > Hans Aberg wrote:
> > >
> > > Here are two interesting links:
> > > http://www.intel.com/products/proces...ium2/index.htm
> > > http://h71028.www7.hp.com/enterprise...0-0-0-121.aspx
> > >
> > > It is a move towards the Itanium, which is big endian, just as the IBM
> > > PowerPC chips, developed by Intel and HP together. So it seems to be both
> > > right business and development moves. One should get even more computing
> > > power.
> > >
> >
> > Itanium is crap - and even intel knows it - they have basically stopepd
> > marketing it
>
> x86 is crap too. It doesn't make sense that Apple would want to switch
> to a processor architecture burdened by such an old instruction set.
> This rumor makes me wonder if Intel is working on a new CPU. For Apple,
> Intel would be a chance to free themselves from chipmakers that don't
> care much about desktop class components. For Intel, Apple would be a
> proof of concept so they can market the new chips to Microsoft.

This switch is not a rumor, Jobs announced the switch today. The
motivating factor is that IBM is not willing or able to have a good
growth path for the G5 chip. Notice that IBM has yet to produce a G5
that could run cool enough to be used in Powerbooks, nor has IBM
produced a G5 that meets or exceeds 3Ghtz in speed. From what I have
read, IBM isn't shedding any tears about this change on Apple's part
because IBM wants to focus its microprocessor division on producing
chips for the Xbox 360 and other video game consoles.

By going with Intel, Apple will hopefully get a faster 64-bit chip for
its desktop Macs and a 64-bit chip for use in future Powerbooks. Nothing
I have seen indicates that Apple will use the Itanium, but than again,
that is a logical assumption. As for Intel not marketing the Itanium,
that's just silly. All the major PC companies are producing some
computers that use the Itanium. I think if Apple keeps Intel under a
tight agreement, this could be a big win for Apple.

Andy Davidson wrote:
> How surprised is everyone that Apple have not picked AMD - I've been wary
> of AMD kit for years, but I'm really impressed with the 64-bit Opteron
> offerints. Very impressed.


$$$$$$$$$$$$$$$$$$$$


Whenever you see a computer add in print or on TV which ends with the
Intel tune and logo, it means that Intel has provided funcing for this
ad, thus reducing the cost to the manufactuer to advertise. AMD doesn't
have such an extensive marketing subsidization programme.

So for Apple, when negotiating with Intel and AMD, looking at all the
costs and subsidies etc, Intel probably came out cheaper than AMD.

Also, at this point in time, Intel is stronger than AMD for laptops.
(this may change in future).

Of course, this means that any Apple AD you may see on TV will now end
with the barf-inducing Intel logo and noise.