At 3:52 AM +0100 1/2/05, Listar wrote:
>From: "Jeff Keller" <jrk_om@xxxxxxxxxxx>
>Subject: [OM] Re: [OT] Is The MAC Dead
>Date: Sat, 1 Jan 2005 11:13:50 -0800
>
>
>Not all microprocessors are used in personal computers. Internet equipment
>uses quite a few (admittedly less than PCs). Compatibility with Microsoft
>adds no value. Power PC chips can generate significant profits.
They do at that. Something like 2/3 of the industrial embedded
computers use PowerPC. The other 1/3 is scattered among a collection
of other chips, including Intel. That said, according to the reports
in the electronics trade magazines, the Mac market is the volume
leader for PowerPC.
Non-industrial embedded devices typically use small processors like
the venerable 8051, 68040, ARM, et al. There are dozens of choices.
None of these are even close to being able to run MacOS 10 or Windows
XP, even if the system did have a disk; most don't.
In much of the the embedded market, Windows would not be considered
for use as the target OS, but is widely used as a development host.
It used to be that UNIX was far more expensive than Windows, but that
is no longer true, and many embedded shops use only UNIX hosts for
development, especially if the target is Linux (which is busily
eating Microsoft's lunch in the server and embedded markets).
>-jeff
>----- Original Message -----
>From: "Julian Davies" <julian_davies@xxxxxxxxxxxxxx>
>Subject: [OM] Re: [OT] Is The MAC Dead
>>
>> No, companies produce what is profitable. However good the power PC is as
>> a chip, it is not in the same league of unit profitability as a Pentium god
>> knows what. Intel will use vastly greater resources to keep it that way,
> > and they have some other areas of performance growth to investigate too.
The PowerPC is a joint development of IBM, Motorola (whose
semiconductor division was just spun off as Freescale), and Apple.
They seem to have the money needed to produce the next generation of
chips on schedule; they have been doing so for years. Nor do they
wish to become hostage to Intel, or to lose their differentiation
from Intel, to become just another seller in a commodity market.
IBM invented the RISC processor, which became the PowerPC of today.
IBM is also a major semiconductor maker, and has been for decades,
supplying their own mainframe computer line.
Intel is actually in a sense trapped by the Pentium. It's getting
harder and harder to eke the last increment of performance out of
that architecture, but if they change to a new and uncluttered
architecture, they will lose the upward compatibility that has kept
the Windows application software base locked down for all these years.
A datapoint: A PowerPC needs only about half the clock rate to match
the computational performance of a Pentium. In other words, a
500-MHz PowerPC is the equal of a 1-GHz Pentium, because the Pentium
architecture is very old and has become encrusted, while the PowerPC
is new and clean.
By contrast, Apple managed to pull off the migration from the
68000-series CPUs to the PowerPC-series quite smoothly. In the
runup to the migration, Apple ran focus group interviews of the
faithful all around the country. I was invited to one of these
groups. We were asked all manner of questions, but the one question
that we were all required to speak directly to was if the 68040
processors needed to be retained on the PowerPC-based Mac
motherboards to come. Opinion was uniform that the 68000s had had
their day, and that it was now OK to leave it behind, existing only
as a MacOS 8 emulation mode. I assume that Apple got this same
answer across the board, because that's exactly what happened: only
the emulation mode remained.
As for porting MacOS to Intel, I don't think it will ever happen.
Apple's business model is that they make most of their money on the
hardware, and so they couldn't survive if someone else sold the
hardware. They would also lose control of the hardware, its
integration with MacOS, and its quality control, thus sacrificing
their competitive advantages of tight hw-sw integration and
reliability. And innovation -- all that R&D costs serious money,
which has to come from somewhere.
Datapoint: One of the first things Steve Jobs did when he returned
to Apple was to buy back the licence of the sole Mac clonemaker,
Power Computing Corp (PCC), for $100 million. But not before PCC
revealed just how insular Apple's hardware engineering had become.
The PCC clones cost half what the real Macs cost, but worked just as
well. (I had one.) This revelation cost the then head of hardware
engineering his job, and his whole department to be "reorganized".
After which, Macs started to use PC parts wherever possible.
Datapoint: In round numbers, the PC market is hundreds of millions,
the Mac market is tens of millions, and the UNIX market is millions.
The rise of Linux and the use of BSD UNIX in MacOS 10 are confounding
this hoary rule-of-thumb, but still the rule is useful.
One can turn the Macs-are-too-expensive argument on its head: Back
in the old days, Macs were about 10% of the market, and cost about
twice what a "comparable" PC cost, however one determined what was
comparable. So, tens of millions of people (~10% of the then PC
market) were sufficiently convinced of the superiority of the Mac
that they were willing to pay double. The current spread isn't
nearly that large, if it exists at all.
Some years ago, when Apple hit bottom, a PC-bigot friend of mine
asked me why I thought Apple would survive. My answer was simply
that 25 million rabid Macophiles wouldn't permit Apple to die.
Joe
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|