On high hopes, misconceptions, and Apple on x86
There’s been a lot of talk about Apple’s recent motion that they’ll be moving to Intel chips. That is to be expected. But in the torrent of opinions, there are a few strong currents that are, shall we say… delusional. I’ve already had to swim against them in several discussions, so I decided to try my hand at some Mac punditry.
“I want MacOS X to run it on my street PC”
You’d be a fool to think that Apple will ever allow that.
Firstly, Apple is a hardware company. They write and sell software too, but that isn’t their primary business. Their profit is in hardware. Allowing their software to run on hardware they didn’t sell is entirely worthless. In fact, it would be harmful as they’d suddenly have to adapt OS X to play well with a sizable portion of possible Wintel hardware combinations, and would then have to support all that software running on foreign hardware. Which brings us to the next point:
The closed hardware platform is a huge competitive advantage in software quality. Microsoft as well as all the i386-based Unices all have a huge headache: there is a bewildering array of hardware they must play well on. It’s impossible to guarantee that things work for everyone. Apple on the other hand can and do test all the software they ship on all of the hardware they sell. The box and the bits you get are guaranteed to play together. Contrast with Windows, where 85% of the catastrophic crashes these days are attributable to thirdparty hardware drivers – but it’s Microsoft who get the flack.
Lastly, if Apple’s operating system could boot on any street PC, can you guess whom that would pit them against as direct competitors? Microsoft. Yeah, that’ll fly.
In fact, I expect that to get OS X to run on a street PC you’ll need some sort of hardware emulation – possibly something relatively benign, like (a part of) the chipset.
But the Osborne effect will kill them!
I don’t think citing the Osborne effect is at all relevant.
Let’s assume for a moment that IBM was intent on not delivering the faster CPUs they promised to Apple. (I’ll get to that shortly.) Extrapolating from there, which scenario is more appealing?
Apple sticks with the PowerPC for a long time. Sales keep growing, but their hardware slowly falls behind on performance relative to the rest of the market. Eventually, their computers are slow enough that sales start dwindling, and Apple are forced to switch. Sales dry up completely.
Apple forsee that this will happen. They decide to switch now. Sales, which were going strong so far, suddenly take a nosedive.
I don’t think there can be any question as to which is the more attractive choice; though “attractive” is obviously an odd choice of word in this instance.
Are those the only options, you may wonder, or is this maybe a false dilemma? Maybe IBM would deliver, providing a third possible scenario that would be better than either of the above?
Just recently Steve Jobs has had to apologize to the Apple community for not being able to deliver on last-year’s promise of a 3-Ghz G5 by mid 2004. IBM promised to make that available, but has not done so.
A lot of people have excused this on the grounds that the move to 90-nanometer manufacturing has proven more difficult than anticipated, but I don’t believe that. PowerPC does not have the absurd complexities of the x86, and 90-nanometer production should be easily in reach for IBM. The cell processor, furthermore, is confidently planned for mass production at 65-nanometer sizes early next year.
This will get more interesting if, as reported on various sites, such as Tom’s Hardware, IBM has been burning the candle at both ends and also will produce a three-way, 3.5-GHz version of the PowerPC for use on Microsoft’s Xbox.
Whether that’s true or not, however, my belief is that IBM chose not to deliver on its commitment to Apple because doing so would have exacerbated the already embarrassing performance gap between its own server products and the higher end Macs. Right now, for example, Apple’s 2-Ghz Xserve is a full generation ahead of IBM’s 1.2-GHz p615, but costs about half as much.
No. Apple had to switch.
Note that iPod sales are not going to magically dry up. And I expect that Apple will dump prices on their current hardware lineup at some point to give waning sales a boost.
Robert X. Cringely posted a theory that with this move Apple is Going for Broke: that this is Intel’s move to break Microsoft. He hypothesizes that Intel, who are disgruntled with Microsoft, want to buy Apple and make OS X available to their OEMs and thus break Mirosoft’s back. But that flat out makes no sense at all:
So Intel buys Apple and works with their OEMs to get products out in the market. The OEMs would love to be able to offer a higher margin product with better reliability than Microsoft.
Sound plausible? Well, see above about the competitive advantage of the closed platform. If OS X were to have to run on more or less the entire range of Wintel hardware, it wouldn’t be more realiable than Microsoft.
There are more false premises where that one came from:
Question 1: What happened to the PowerPC’s supposed performance advantage over Intel?
Nothing! The high-end IBM chips were much faster than Intel’s offerings at the time. But that was then; ever since, IBM has been hemming and hawing but not giving Apple any faster chips. Nor are they planning to.
Apple did not decide to switch because the PowerPC is slow. They decided to switch because it’s not getting any faster.
Question 4: Why announce this chip swap a year before it will even begin for customers?
Here, Cringely argues that the Osborne effect would totally kill Apple, if they weren’t actually planning something completely different (
the Apple/Intel merger). But see above: this makes no sense. Apple is in a position of strength right now and they have a supporting business that will help carry them through the Osborne effect.
Now consider the other option. Suppose we are in an alternate universe, and Apple’s x86 hardware is ready to ship now.
There’s no native software for it. A lot of software runs under emulation, but it runs slower than on PowerPC. The effect on the PowerPC line of hardware is be the same: a dramatic drop. But the x86 hardware doesn’t sell any better either! In fact, it’s possible that in this alternate universe, sales of PowerPC Macs crash even more dramatically than they do in ours, while sales of x86 Macs plod along for a long time.
What does Apple stand to win in that scenario? Not much, if anything.
On top of all that, I somehow can’t imagine people to get all excited about Macintoshes built by Intel. I believe customers would jump ship on the platform in droves if such a merger came to pass.
Cringely’s scenario is built on faulty assumptions and full of holes. It would never work that way.
Nothing. It’s likely there is a bigger master plan here. I’m not going to touch the predictions game with a 10-foot pole, however. All I can tell is that everyone who has gotten into it so far would probably have done well not to. I know enough to heed that sort of warning.