By John Gruber
Halide Mark II: The Best Pro Camera for iPhone
All the Great Men in the history of business, they founded their own companies, and shaped them in their image. Henry Ford. Ray Kroc. Bill Gates. Walt Disney.
Robert X. Cringely has, for several years, been hung up on the idea that Steve Jobs wants to be the CEO of Disney. Cringely is far from the only pundit to float this idea, and on the surface it makes a wee bit of sense: Disney has not been doing very well for at least a decade, and their most consistently successful endeavors in that time have been the films they’ve released from Pixar.
But Jobs doesn’t want to take over Disney — he wants to turn Apple into a Disney-esque company. Not in terms of the products it produces, of course, but in terms of cultural impact, brand power, and worldwide financial success. Jobs thinks of himself as a peer to Walt Disney, a visionary who creates something of his own. Nothing at all like, say, Michael Eisner — an executive who simply comes in and runs a company founded decades earlier.
Disney is always going to be Walt Disney’s company.
And Apple — his decade-long exile notwithstanding — is always going to be Steve Jobs’s company.
And so Cringely’s column this week, wherein he speculates that Intel is going to buy Apple, overreaches by far. In any hypothetical scenario where Apple gets bought by a larger company, the first question you need to answer is “Where do you put Steve Jobs?” He is not a second-in-command sort of guy. And so Cringely files him away thusly:
This scenario works well for everyone except Microsoft. If Intel was able to own the Mac OS and make it available to all the OEMs, it could break the back of Microsoft. And if they tuned the OS to take advantage of unique features that only Intel had, they would put AMD back in the box, too. Apple could return Intel to its traditional role of being where all the value was in the PC world. And Apple/Intel could easily extend this to the consumer electronics world. How much would it cost Intel to buy Apple? Not much. And if they paid in stock it would cost nothing at all since investors would drive shares through the roof on a huge swell of user enthusiasm.
That’s the story as I see it unfolding. Steve Jobs finally beats Bill Gates. And with the sale of Apple to Intel, Steve accepts the position of CEO of the Pixar/Disney/Sony Media Company.
That’s not to say it isn’t a good read, however. The basic idea is sound and reasonable — Apple and Intel suddenly seem to be natural allies, pitted, albeit somewhat obliquely, against Microsoft. The relationship between Microsoft and Intel has been strained in recent years, and Microsoft’s decision to switch to PowerPC processors for the Xbox 360 can only be seen as making matters even worse.
Apple matters to Intel not because of its market share — Apple could double or triple its PC hardware market share with its Intel-based Macs and it would barely make a difference in Intel’s overall share of the CPU market.
What’s appealing about Apple to Intel is that Apple makes innovative products. They push the limits, and are not afraid to introduce new technology first. The commodity PC hardware business has been one of incremental change, with PC makers clinging to legacy buses and ports for as long as they can.
In short, Intel would like to see a bit less evolution and a bit more revolution in the way its chips are being used. Apple is that type of company.
For years Intel has been making cool prototype “PCs of the future”, but no one ever turns their ideas into actual products, instead making new revisions of the same PC designs, only faster. There is a herd mentality amongst PC makers, much like in any commodity business. There is nothing herd-like whatsoever to Apple’s mentality. With Apple, Intel has gained a partner that is unafraid to break new ground.
What’s appealing about Intel to Apple is that they solve both of Apple’s long-term CPU problems:
The PowerPC platform has often lagged in performance. Recall, most notoriously, when Motorola’s G4 production was maxed out at 500 MHz for over a year.
Apple has occasionally had problems getting enough PowerPC chips to meet the demand for its products. There have been times when Apple couldn’t make computers fast enough to meet demand because they couldn’t get enough CPUs.
Problem #2 is one of the major differences between Intel and AMD. AMD’s processors may be faster than Intel’s — that’s a race that will likely continue going back and forth — but Intel, inarguably, has greater production capabilities than any other semiconductor manufacturer. Intel makes more CPUs than the rest of the industry combined. No other CPU partner can make it less likely that Apple will ever again face a CPU shortage.
(Interpolation on the “Why not AMD?” Question: I’ve seen numerous arguments that Apple has somehow made a terrible mistake by choosing Intel rather than AMD, mostly on the grounds that AMD processors currently have a slight edge over Intel’s, performance-wise. That’s nonsense. If you assume that Apple negotiated with AMD as well — and it’s hard to imagine they wouldn’t have — Intel must have offered a better deal. Better prices, better roadmap, better power consumption — we don’t know what, because only Apple’s and Intel’s executives know the details of the deal, but it must have been better or more appealing. And most importantly, a few years down the road, once the Mac OS X software transition to the x86 instruction set is complete, Apple can switch to (or add to their mix) AMD processors just as easily as they currently use PowerPC processors from both Freescale (née Motorola) and IBM.)
We don’t yet know exactly what Apple’s product plans are for their upcoming Intel-based systems, but it seems likely that Intel will be supplying more than just the processors. What I expect is that internally, Intel-based Macs will be largely designed and engineered by Intel. Internally, they’ll be more similar than different to standard PCs. Consider Phil Schiller’s statement that it will likely be possible, albeit not sanctioned by Apple, to boot these Macs using Windows — this wouldn’t be feasible if Apple weren’t planning to use a similar internal architecture to standard PCs.
There is a certain contingent of Mac nerd for whom this is depressing. These are people who get off on the idea that Apple should blaze its own trails, and who hold out hope that Apple would eventually do something that left the rest of the PC industry in the dust. There were flashes of this potential with each new leap of the PowerPC — when the first wave of PowerMacs hit in the 90’s; when the first PowerPC PowerBooks arrived; when the G4 debuted; and most recently, with the debut of the PowerMac G5. With these surges, Apple could occasionally leapfrog the rest of the PC industry.
The problem is, they never stayed there. Steadily but surely, every time Apple was able to surge ahead performance-wise, Intel and AMD would catch up and then overtake the lead. It was an enticing thought that Apple might one day surge ahead and stay there, but it never amounted to more than that: a nice thought.
By switching to a mostly industry-standard Intel architecture, Apple obviously has no chance of ever surpassing the PC industry in performance. But the other side of that argument is equally true: they’ll never again fall behind, either.
Many people, including Cringely, have asked what this news means regarding Apple’s previous touting of the PowerPC versus Intel CPUs. Cringely writes:
Apple loved to pull Phil Schiller onstage to do side-by-side speed tests showing how much faster in real life the G4s and G5s were than their Pentium equivalents. Was that so much BS? Did Apple not really mean it? And why was the question totally ignored in this week’s presentation?
It was never bullshit. What it was, was selective timing. Apple didn’t perform on-stage benchmarks during every keynote — they did so only during the periods when their computers had surged ahead in the performance war. That’s not say they’ve now fallen dramatically behind — it’s just that those demos were only worth performing when they were dramatically ahead, and those opportunities were few and far between.
If, however, you were left with the impression that Power Macs had remained as far ahead of the latest-and-greatest from Dell as they were the last time Schiller did his Photoshop race on-stage, well, Apple didn’t mind. If you want to call that “bullshit”, go ahead, but it’s rather unrealistic to expect Apple to bring your attention to benchmarks that don’t make its products look good.
As for why it was ignored during the presentation, that’s quite obvious. Until they start shipping Intel-based Macs, Apple is walking a tightrope: building demand for the upcoming new machines, but while still selling as many existing PowerPC-based Macs as they can.
Second, Jobs’s point of emphasis in the keynote was the metric of power-for-performance, not raw performance. He emphasized that Apple’s decision to switch was based on a desire to make better computers, not just faster ones.
In a nut: if you can’t beat them, join them.
Here’s another way to look at this deal: what Darwin is to Mac OS X, Intel will supply for Mac hardware.
What I mean by this is that by basing Mac OS X around the open-source Unix-derived Darwin core, Apple essentially gave up on the idea of creating their own OS software from scratch. Is Unix ideal? No. Is Unix good enough? Yes. And it allowed Apple to move forward and put their effort into the parts of the OS that separate them from everyone else: Cocoa, Carbon, QuickTime, and their own applications.
With this Intel partnership, Apple is doing the same thing for their hardware. No more going it alone, designing the entire kit. With a core architecture designed by Intel, Apple will be free to concentrate their own engineering on what they do best: the user experience. Not just how the case looks, but how it feels. In the end, it’s the product that matters, not the underlying technology. Mac OS X established that for their operating system, and this Intel deal will do the same for the hardware.
This doesn’t mean Mac hardware will simply be “a PC with a pretty case” any more than that Mac OS X is merely “Unix with a pretty UI”. It just means that both Apple and Intel get to concentrate on the things that they do best.
So Apple and Intel are very complementary partners. Other than the software transition and sales hit Apple is going to take in the interim, it ought to work out well for both companies.
And, if it goes really well, it’s easy to see how this could perhaps lead to a significant uptick in the market share of PCs running Mac OS X. For one thing, it’s easy to see how this partnership with Intel could open doors to the corporate enterprise market that have remained closed to Apple. Is it likely that Apple will eventually start licensing Mac OS X to other PC makers? I say no. But I’d be a fool to say it isn’t at least a little more likely now than it was a week ago. And, maybe it’s now a lot more likely.
So Cringely is right, in that if you start thinking big thoughts, this could be the start of a very powerful alliance, in both direct and oblique confrontation with Microsoft. Not just in terms of the existing PC market, but in the escalating fusion of the PC and consumer electronics industries. Or as the subtitle to Cringely’s column states, “Apple’s Decision to Use Intel Processors Is Nothing Less Than an Attempt to Dethrone Microsoft. Really.”
If this week’s news proved anything, it’s that Apple is unafraid to act boldly. But none of this means it makes sense for Intel to buy Apple. What advantage would be gained by merging rather than developing a symbiotic partnership? A buyout would add nothing of benefit, but would incur all sorts of other problems. For one thing, it would obviously antagonize all of Intel’s existing customers for CPUs. And as stated at the outset, it defies belief to think that Steve Jobs wants to cede control of Apple Computer.
Cringely, as he often does, started with a good idea but took it one click too far, turning the pundit dial to 11 when 10 would have done just fine, and remained within the realm of plausibility.